Daily Papers

To render each generated token in real-time for users, the Large Language Model (LLM) server generates tokens one by one and streams each token (or group of a few tokens) through the network to the user right after generation, which we refer to as LLM token streaming. However, under unstable network conditions, the LLM token streaming experience could suffer greatly from stalls since one packet loss could block the rendering of later tokens even if the packets containing them arrive on time. With a measurement study, we show that current applications suffer from increased stalls under unstable networks. For this emerging token streaming problem in LLM Chatbots that differs from previous multimedia and text applications, we propose a novel transmission scheme, called Eloquent, which puts newly generated tokens as well as currently unacknowledged tokens in the next outgoing packet. This ensures that each packet contains some new tokens and, in the meantime, is independently rendered when received, avoiding the aforementioned stalls caused by missing packets. Through simulation under various networks, we show Eloquent reduces stall ratio (proportion of token rendering wait time) by 71.0% compared to the retransmission method commonly used by real chatbot applications and by 31.6% compared to the baseline packet duplication scheme. By tailoring Eloquent to fit the token-by-token generation of LLM, we enable the Chatbots to respond like an eloquent speaker for users to better enjoy pervasive AI.

Recent advancements in speech-to-speech dialogue systems leverage LLMs for multimodal interactions, yet they remain hindered by fine-tuning requirements, high computational overhead, and text-speech misalignment. Existing speech-enabled LLMs often degrade conversational quality by modifying the LLM, thereby compromising its linguistic capabilities. In contrast, we propose LLMVoX, a lightweight 30M-parameter, LLM-agnostic, autoregressive streaming TTS system that generates high-quality speech with low latency, while fully preserving the capabilities of the base LLM. Our approach achieves a significantly lower Word Error Rate compared to speech-enabled LLMs, while operating at comparable latency and UTMOS score. By decoupling speech synthesis from LLM processing via a multi-queue token streaming system, LLMVoX supports seamless, infinite-length dialogues. Its plug-and-play design also facilitates extension to various tasks with different backbones. Furthermore, LLMVoX generalizes to new languages with only dataset adaptation, attaining a low Character Error Rate on an Arabic speech task. Additionally, we have integrated LLMVoX with a Vision-Language Model to create an omni-model with speech, text, and vision capabilities, without requiring additional multimodal training. Our code base and project page is available at https://mbzuai-oryx.github.io/LLMVoX .

Speculative decoding accelerates LLM inference by using a draft model to look ahead, but gains are capped by the cost of autoregressive draft generation: increasing draft size elevates acceptance rates but introduces additional latency overhead exacerbating the speed-accuracy tradeoff. Prior methods (Medusa, Hydra, EAGLE) partially reduce draft cost but either degrade acceptance or introduce overheads that limit scaling. We present Mirror Speculative Decoding (Mirror-SD), an inference algorithm that breaks the latency-acceptance tradeoff. Mirror-SD launches branch-complete rollouts from early-exit signals in parallel with the target model's suffix and explicitly maps computation across heterogeneous accelerators (GPU and NPU) to exploit cross-device parallelism. The draft speculates forward continuations for the target to verify, while the target simultaneously speculates correction paths for the draft, converting speculation into two complementary execution pipelines. To further cut draft latency without weakening acceptance semantics, we add speculative streaming so the draft emits multiple tokens per step. This dual strategy of parallel heterogeneous execution plus multi-token speculative streaming pushes speculative decoding toward its ideal regime of high acceptance with low overhead. On SpecBench with server-scale models from 14B to 66B parameters, Mirror-SD delivers consistent end-to-end gains, achieving 2.8x-5.8x wall-time speedups across diverse tasks and a 30% average relative improvement over the strongest baseline, EAGLE3.

The integration of Large Language Models (LLMs) into applications ranging from interactive chatbots to multi-agent systems has introduced a wide spectrum of service-level objectives (SLOs) for responsiveness. These include latency-sensitive requests emphasizing per-token latency in streaming chat, deadline-sensitive requests requiring rapid full responses to trigger external tools, and compound requests with evolving dependencies across multiple LLM calls. Despite-or perhaps, because of-this workload diversity and unpredictable request information (e.g., response lengths and dependencies), existing request schedulers have focused on aggregate performance, unable to ensure application-level SLO needs. This paper presents JITServe, the first SLO-aware LLM serving system designed to maximize service goodput (e.g., the number of tokens meeting request SLOs) across diverse workloads. JITServe novelly schedules requests using imprecise request information and gradually relaxes this conservatism by refining request information estimates as generation progresses. It applies a grouped margin goodput maximization algorithm to allocate just enough serving bandwidth to satisfy each request's SLO just-in-time (JIT), maximizing residual capacity for others, while deciding the composition of requests in a batch to maximize efficiency and goodput with provable guarantees. Our evaluation across diverse realistic workloads, including chat, deep research, and agentic pipelines, shows that JITServe improves service goodput by 1.4x-6.3x, alternatively achieving 28.5%-83.2% resource savings, compared to state-of-the-art designs.

Despite advances in 3D scene understanding, existing 3D Large Multimodal Models operate in offline settings, requiring complete scene observations or predefined video clips. In this paper, we present an online 3D vision-language model that enables real-time spatial understanding from streaming video. Our approach adopts an autoregressive streaming control modeling based on the LLM's next-token prediction objective to learn when to respond, and employs a lightweight Visual-Spatial Feature Integration (VSFI) module to incrementally inject temporally aligned geometry priors into the visual stream. To alleviate long-context decoding overhead, we propose a plug-and-play Geometry-Adaptive Voxel Compression (GAVC) module for efficient visual token compression. To address the scarcity of streaming 3D-language data, we further develop a scalable data generation pipeline that curates over 1M online spatio-temporal 3D QA pairs and establishes a comprehensive benchmark spanning 29 tasks. Extensive experiments show that our approach significantly outperforms both proprietary and open-source models across online and offline 3D spatial understanding, reasoning, and grounding tasks. The project page is available at https://stream3d-vlm.github.io/

As Large Language Models (LLMs) become increasingly prevalent in various domains, their ability to process inputs of any length and maintain a degree of memory becomes essential. However, the one-off input of overly long texts is limited, as studies have shown that when input lengths exceed the LLMs' pre-trained text length, there is a dramatic decline in text generation capabilities. Moreover, simply extending the length of pre-training texts is impractical due to the difficulty in obtaining long text data and the substantial memory consumption costs this would entail for LLMs. Recent efforts have employed streaming inputs to alleviate the pressure of excessively long text inputs, but this approach can significantly impair the model's long-term memory capabilities. Motivated by this challenge, we introduce Streaming Infinite Retentive LLM (SirLLM), which allows LLMs to maintain longer memory during infinite-length dialogues without the need for fine-tuning. SirLLM utilizes the Token Entropy metric and a memory decay mechanism to filter key phrases, endowing LLMs with both long-lasting and flexible memory. We designed three distinct tasks and constructed three datasets to measure the effectiveness of SirLLM from various angles: (1) DailyDialog; (2) Grocery Shopping; (3) Rock-Paper-Scissors. Our experimental results robustly demonstrate that SirLLM can achieve stable and significant improvements across different LLMs and tasks, compellingly proving its effectiveness. When having a coversation, "A sir could forget himself," but SirLLM never does! Our code is publicly available at https://github.com/Zoeyyao27/SirLLM

We introduce Guard Vector, a safety task vector computed as the parameter difference between a guardrail model (Guard Model) and a same-architecture pretrained language model. Composing this vector with a target language model yields a Target Guard Model (TGM). We then adapt TGM with a streaming-aware approach that combines prefix-based training and evaluation with a classifier that produces a single-token output. With this composition alone, TGM improves classification quality over established Guard Models across standard safety suites and enables language extensibility to Chinese, Japanese, and Korean, requiring neither additional training nor target language labels. It also demonstrates model portability across two widely used public guardrail backbones, Llama and Gemma. With prefix SFT (supervised fine-tuning), TGM preserves classification quality under streaming by aligning the behavior between prefix inputs and full-text inputs. The single-token output design increases throughput and reduces latency. Together, these components reduce data and compute requirements while promoting streaming-aware evaluation practices, thereby contributing to a more responsible AI ecosystem.

Context retrieval systems for LLM inference face a critical challenge: high retrieval latency creates a fundamental tension between waiting for complete context (poor time-to-first-token) and proceeding without it (reduced quality). Streaming context incrementally--overlapping retrieval with inference--can mitigate this latency, but doing so with concurrent requests introduces new challenges: requests contend for GPU compute and memory, and scheduling must adapt to dynamic context arrivals. We present Stream2LLM, a streaming-aware LLM serving system for concurrent prefill-decode disaggregated deployments. Stream2LLM introduces adaptive scheduling and preemption for two distinct retrieval patterns: append-mode (progressive context accumulation) and update-mode (iterative refinement with cache invalidation). It decouples scheduling decisions from resource acquisition, enabling flexible preemption strategies guided by hardware-specific cost models, and uses longest common prefix matching to minimize redundant computation when input changes dynamically. To evaluate Stream2LLM, we collect two large-scale, real-world streaming workloads based on web crawling and approximate nearest neighbor search. Our evaluation demonstrates that streaming architecture delivers up to 11x TTFT improvements, with cost-aware scheduling providing critical benefits under memory pressure, all while maintaining throughput parity with non-streaming baselines. Code: https://github.com/rajveerb/stream2llm/tree/mlsys_artifact

Streaming video large language models (LLMs) are increasingly used for real-time multimodal tasks such as video captioning, question answering, conversational agents, and augmented reality. However, these models face fundamental memory and computational challenges because their key-value (KV) caches grow substantially with continuous streaming video input. This process requires an iterative prefill stage, which is a unique feature of streaming video LLMs. Due to its iterative prefill stage, it suffers from significant limitations, including extensive computation, substantial data transfer, and degradation in accuracy. Crucially, this issue is exacerbated for edge deployment, which is the primary target for these models. In this work, we propose V-Rex, the first software-hardware co-designed accelerator that comprehensively addresses both algorithmic and hardware bottlenecks in streaming video LLM inference. At its core, V-Rex introduces ReSV, a training-free dynamic KV cache retrieval algorithm. ReSV exploits temporal and spatial similarity-based token clustering to reduce excessive KV cache memory across video frames. To fully realize these algorithmic benefits, V-Rex offers a compact, low-latency hardware accelerator with a dynamic KV cache retrieval engine (DRE), featuring bit-level and early-exit based computing units. V-Rex achieves unprecedented real-time of 3.9-8.3 FPS and energy-efficient streaming video LLM inference on edge deployment with negligible accuracy loss. While DRE only accounts for 2.2% power and 2.0% area, the system delivers 1.9-19.7x speedup and 3.1-18.5x energy efficiency improvements over AGX Orin GPU. This work is the first to comprehensively tackle KV cache retrieval across algorithms and hardware, enabling real-time streaming video LLM inference on resource-constrained edge devices.

Many LLM tasks are performed in large batches or even offline, and the performance indictor for which is throughput. These tasks usually show the characteristic of prefix sharing, where different prompt input can partially show the common prefix. However, the existing LLM inference engines tend to optimize the streaming requests and show limitations of supporting the large batched tasks with the prefix sharing characteristic. The existing solutions use the LRU-based cache to reuse the KV context of common prefix. The KV context that is about to be reused may prematurely be evicted with the implicit cache management. Even if not evicted, the lifetime of the shared KV context is extended since requests sharing the same context are not scheduled together, resulting in larger memory usage. These streaming oriented systems schedule the requests in the first-come-first-serve or similar order. As a result, the requests with larger ratio of decoding steps may be scheduled too late to be able to mix with the prefill chunks to increase the hardware utilization. Besides, the token and request number based batching can limit the size of token-batch, which keeps the GPU from saturating for the iterations dominated by decoding tokens. We propose BatchLLM to address the above problems. BatchLLM explicitly identifies the common prefixes globally. The requests sharing the same prefix will be scheduled together to reuse the KV context the best, which also shrinks the lifetime of common KV memory. BatchLLM reorders the requests and schedules the requests with larger ratio of decoding first to better mix the decoding tokens with the latter prefill chunks and applies memory-centric token batching to enlarge the token-batch sizes, which helps to increase the GPU utilization. Extensive evaluation shows that BatchLLM outperforms vLLM by 1.1x to 2x on a set of microbenchmarks and two typical industry workloads.

Streaming Video Large Language Models (VideoLLMs) have demonstrated impressive performance across various video understanding tasks, but they face significant challenges in real-time deployment due to the high computational cost of processing dense visual tokens from continuous video streams. In streaming video scenarios, the primary bottleneck lies in the Vision Transformer (ViT) encoding stage, where redundant processing of temporally similar frames leads to inefficiency. Additionally, inflated token sequences during LLM pre-filling further exacerbate latency and memory overhead. To address these challenges, we propose Streaming Token Compression (STC), a plug-and-play hierarchical framework that seamlessly integrates into existing streaming VideoLLMs, optimizing both ViT encoding and LLM pre-filling stages to accelerate processing. STC introduces two token-level accelerators: STC-Cacher, which reduces ViT encoding overhead by caching and reusing features from temporally similar frames, and STC-Pruner, which compresses the visual token sequence before it enters the LLM, preserving only the most salient tokens based on both spatial and temporal relevance. Extensive experiments on four baseline streaming VideoLLMs across five benchmarks demonstrate that STC outperforms other compression methods. Notably, STC retains up to 99\% of accuracy on the ReKV framework while reducing ViT encoding latency and LLM pre-filling latency by 24.5\% and 45.3\%.

Speculative decoding is a prominent technique to speed up the inference of a large target language model based on predictions of an auxiliary draft model. While effective, in application-specific settings, it often involves fine-tuning both draft and target models to achieve high acceptance rates. As the number of downstream tasks grows, these draft models add significant complexity to inference systems. We propose Speculative Streaming, a single-model speculative decoding method that fuses drafting into the target model by changing the fine-tuning objective from next token prediction to future n-gram prediction. Speculative Streaming speeds up decoding by 1.8 - 3.1X in a diverse set of tasks, such as Summarization, Structured Queries, and Meaning Representation, without sacrificing generation quality. Additionally, Speculative Streaming is parameter-efficient. It achieves on-par/higher speed-ups than Medusa-style architectures while using ~10000X fewer extra parameters, making it well-suited for resource-constrained devices.

Video Large Language Models (Video-LLMs) excel at understanding videos in-context, provided they have full access to the video when answering queries. However, these models face challenges in streaming scenarios where hour-long videos must be processed online, and questions need timely responses. In this work, we propose a training-free approach compatible with standard Video-LLMs, leveraging three key concepts: 1) LLM-informed selection of visual tokens to identify those that the LLM has attended to and contributed to its understanding of each short clip. Our attention-based selection allows us to discard up to ~95% of unimportant visual tokens with minimal performance loss; 2) Recurrent processing of past selected tokens to generate temporally coherent understanding of each processed clip; 3) Caption-based question answering for lightweight and accurate responses. Our method achieves state-of-the-art performance on streaming video benchmarks, striking a balance between efficiency and effectiveness.

Though safety alignment has been applied to most large language models (LLMs), LLM service providers generally deploy a subsequent moderation as the external safety guardrail in real-world products. Existing moderators mainly practice a conventional full detection, which determines the harmfulness based on the complete LLM output, causing high service latency. Recent works pay more attention to partial detection where moderators oversee the generation midway and early stop the output if harmfulness is detected, but they directly apply moderators trained with the full detection paradigm to incomplete outputs, introducing a training-inference gap that lowers the performance. In this paper, we explore how to form a data-and-model solution that natively supports partial detection. For the data, we construct FineHarm, a dataset consisting of 29K prompt-response pairs with fine-grained annotations to provide reasonable supervision for token-level training. Then, we propose the streaming content monitor, which is trained with dual supervision of response- and token-level labels and can follow the output stream of LLM to make a timely judgment of harmfulness. Experiments show that SCM gains 0.95+ in macro F1 score that is comparable to full detection, by only seeing the first 18% of tokens in responses on average. Moreover, the SCM can serve as a pseudo-harmfulness annotator for improving safety alignment and lead to a higher harmlessness score than DPO.

The advent of large language models (LLMs) has transformed text-based services, enabling capabilities ranging from real-time translation to AI-driven chatbots. However, existing serving systems primarily focus on optimizing server-side aggregate metrics like token generation throughput, ignoring individual user experience with streamed text. As a result, under high and/or bursty load, a significant number of users can receive unfavorable service quality or poor Quality-of-Experience (QoE). In this paper, we first formally define QoE of text streaming services, where text is delivered incrementally and interactively to users, by considering the end-to-end token delivery process throughout the entire interaction with the user. Thereafter, we propose Andes, a QoE-aware serving system that enhances user experience for LLM-enabled text streaming services. At its core, Andes strategically allocates contended GPU resources among multiple requests over time to optimize their QoE. Our evaluations demonstrate that, compared to the state-of-the-art LLM serving systems like vLLM, Andes improves the average QoE by up to 3.2times under high request rate, or alternatively, it attains up to 1.6times higher request rate while preserving high QoE.

While large language models (LLMs) have revolutionized text-to-speech (TTS) synthesis through discrete tokenization paradigms, current architectures exhibit fundamental tensions between three critical dimensions: 1) irreversible loss of acoustic characteristics caused by quantization of speech prompts; 2) stringent dependence on precisely aligned prompt speech-text pairs that limit real-world deployment; and 3) catastrophic forgetting of the LLM's native text comprehension during optimization for speech token generation. To address these challenges, we propose an LLM-based text-to-speech Generation approach Optimized via a novel dual-branch ArchiTecture (GOAT-TTS). Our framework introduces two key innovations: (1) The modality-alignment branch combines a speech encoder and projector to capture continuous acoustic embeddings, enabling bidirectional correlation between paralinguistic features (language, timbre, emotion) and semantic text representations without transcript dependency; (2) The speech-generation branch employs modular fine-tuning on top-k layers of an LLM for speech token prediction while freezing the bottom-k layers to preserve foundational linguistic knowledge. Moreover, multi-token prediction is introduced to support real-time streaming TTS synthesis. Experimental results demonstrate that our GOAT-TTS achieves performance comparable to state-of-the-art TTS models while validating the efficacy of synthesized dialect speech data.

Large language models (LLMs) have demonstrated remarkable capabilities in chain of thought (CoT) reasoning. However, the current LLM reasoning paradigm initiates thinking only after the entire input is available, which introduces unnecessary latency and weakens attention to earlier information in dynamic scenarios. Inspired by human cognition of thinking while reading, we first design a \textbf{streaming thinking} paradigm for LLMs, where reasoning unfolds in the order of input and further adjusts its depth once reading is complete. We instantiate this paradigm with StreamingThinker, a framework that enables LLMs to think while reading through the integration of streaming CoT generation, streaming-constraint training, and streaming parallel inference. Specifically, StreamingThinker employs streaming reasoning units with quality control for CoT generation, enforces order-preserving reasoning through streaming attention masks and position encoding, and leverages parallel KV caches that decouple input encoding from reasoning generation, thereby ensuring alignment and enabling true concurrency. We evaluate StreamingThinker on the Qwen3 model family across math reasoning, logical reasoning, and context-based QA reasoning tasks. Experimental results show that the StreamingThinker preserves performance comparable to batch thinking, while yielding an 80\% reduction in token waiting before the onset of reasoning and a more than 60\% reduction in time-level latency for producing the final answer, demonstrating the effectiveness of the streaming paradigm for LLM reasoning. Code is publicly available at https://github.com/EIT-NLP/StreamingLLM/tree/main/StreamingThinker.

Recent advances have demonstrated the potential of decoderonly large language models (LLMs) for automatic speech recognition (ASR). However, enabling streaming recognition within this framework remains a challenge. In this work, we propose a novel streaming ASR approach that integrates a read/write policy network with monotonic chunkwise attention (MoChA) to dynamically segment speech embeddings. These segments are interleaved with label sequences during training, enabling seamless integration with the LLM. During inference, the audio stream is buffered until the MoChA module triggers a read signal, at which point the buffered segment together with the previous token is fed into the LLM for the next token prediction. We also introduce a minimal-latency training objective to guide the policy network toward accurate segmentation boundaries. Furthermore, we adopt a joint training strategy in which a non-streaming LLM-ASR model and our streaming model share parameters. Experiments on the AISHELL-1 and AISHELL-2 Mandarin benchmarks demonstrate that our method consistently outperforms recent streaming ASR baselines, achieving character error rates of 5.1% and 5.5%, respectively. The latency optimization results in a 62.5% reduction in average token generation delay with negligible impact on recognition accuracy

Long streaming video QA remains challenging due to growing visual tokens and limited reasoning length of large language models (LLMs). KV-caching stores the Key-Value (KV) of the historical tokens via LLM prefill and enables more efficient streaming QA. However, existing methods cache every one or two frames, causing redundant memory usage and losing fine-grained spatial details within frame or temporal contexts across frames. This paper proposes MuKV, a method that features a multi-grained KV cache compression module and a semi-hierarchical retrieval approach to improve both efficiency and accuracy for long streaming VideoQA. For the offline KV cache, MuKV extracts visual representations at patch-, frame-, and segment-levels. The multiple levels of granularity preserve both local cues and global temporal context, while maintaining efficiency with a dual signal token compression mechanism guided by self-attention and frequency. For online QA, MuKV designs a semi-hierarchical retrieval method to retrieve relevant KV caches for answer generation. Experiments on long-streaming VideoQA benchmarks show that MuKV significantly improves answer accuracy, without sacrificing memory and online QA efficiency. Moreover, our compression mechanism alone brings consistent benefits across answer accuracy, memory, and QA efficiency over baselines, showcasing highly effective contribution.

Deploying Large Language Models (LLMs) in streaming applications such as multi-round dialogue, where long interactions are expected, is urgently needed but poses two major challenges. Firstly, during the decoding stage, caching previous tokens' Key and Value states (KV) consumes extensive memory. Secondly, popular LLMs cannot generalize to longer texts than the training sequence length. Window attention, where only the most recent KVs are cached, is a natural approach -- but we show that it fails when the text length surpasses the cache size. We observe an interesting phenomenon, namely attention sink, that keeping the KV of initial tokens will largely recover the performance of window attention. In this paper, we first demonstrate that the emergence of attention sink is due to the strong attention scores towards initial tokens as a ``sink'' even if they are not semantically important. Based on the above analysis, we introduce StreamingLLM, an efficient framework that enables LLMs trained with a finite length attention window to generalize to infinite sequence lengths without any fine-tuning. We show that StreamingLLM can enable Llama-2, MPT, Falcon, and Pythia to perform stable and efficient language modeling with up to 4 million tokens and more. In addition, we discover that adding a placeholder token as a dedicated attention sink during pre-training can further improve streaming deployment. In streaming settings, StreamingLLM outperforms the sliding window recomputation baseline by up to 22.2x speedup. Code and datasets are provided at https://github.com/mit-han-lab/streaming-llm.

Streaming TTS that receives streaming text is essential for interactive systems, yet this scheme faces two major challenges: unnatural prosody due to missing lookahead and long-form collapse due to unbounded context. We propose a prosodic-boundary-aware post-training strategy, adapting a pretrained LLM-based TTS model using weakly time-aligned data. Specifically, the model is adapted to learn early stopping at specified content boundaries when provided with limited future text. During inference, a sliding-window prompt carries forward previous text and speech tokens, ensuring bounded context and seamless concatenation. Evaluations show our method outperforms CosyVoice-Style interleaved baseline in both short and long-form scenarios. In long-text synthesis, especially, it achieves a 66.2% absolute reduction in word error rate (from 71.0% to 4.8%) and increases speaker and emotion similarity by 16.1% and 1.5% relatively, offering a robust solution for streaming TTS with incremental text.

Deploying long-context large language models (LLMs) is essential but poses significant computational and memory challenges. Caching all Key and Value (KV) states across all attention heads consumes substantial memory. Existing KV cache pruning methods either damage the long-context capabilities of LLMs or offer only limited efficiency improvements. In this paper, we identify that only a fraction of attention heads, a.k.a, Retrieval Heads, are critical for processing long contexts and require full attention across all tokens. In contrast, all other heads, which primarily focus on recent tokens and attention sinks--referred to as Streaming Heads--do not require full attention. Based on this insight, we introduce DuoAttention, a framework that only applies a full KV cache to retrieval heads while using a light-weight, constant-length KV cache for streaming heads, which reduces both LLM's decoding and pre-filling memory and latency without compromising its long-context abilities. DuoAttention uses a lightweight, optimization-based algorithm with synthetic data to identify retrieval heads accurately. Our method significantly reduces long-context inference memory by up to 2.55x for MHA and 1.67x for GQA models while speeding up decoding by up to 2.18x and 1.50x and accelerating pre-filling by up to 1.73x and 1.63x for MHA and GQA models, respectively, with minimal accuracy loss compared to full attention. Notably, combined with quantization, DuoAttention enables Llama-3-8B decoding with 3.3 million context length on a single A100 GPU. Code is provided in https://github.com/mit-han-lab/duo-attention.

Full-duplex speech interaction, as the most natural and intuitive mode of human communication, is driving artificial intelligence toward more human-like conversational systems. Traditional cascaded speech processing pipelines suffer from critical limitations, including accumulated latency, information loss, and error propagation across modules. To address these issues, recent efforts focus on the end-to-end audio large language models (LLMs) like GPT-4o, which primarily unify speech understanding and generation task. However, most of these models are inherently half-duplex, and rely on a suite of separate, task-specific front-end components, such as voice activity detection (VAD) and turn-taking detection (TD). In our development of speech assistant, we observed that optimizing the speech front-end is equally crucial as advancing the back-end unified model for achieving seamless, responsive interactions. To bridge this gap, we propose the first unified audio front-end LLM (UAF) tailored for full-duplex speech systems. Our model reformulates diverse audio front-end tasks into a single auto-regressive sequence prediction problem, including VAD, TD, speaker recognition (SR), automatic speech recognition (ASR) and question answer (QA). It takes streaming fixed-duration audio chunk (e.g., 600 ms) as input, leverages a reference audio prompt to anchor the target speaker at the beginning, and regressively generates discrete tokens encoding both semantic content and system-level state controls (e.g., interruption signals). Experiments demonstrate that our model achieves leading performance across multiple audio front-end tasks and significantly enhances response latency and interruption accuracy in real-world interaction scenarios.

Scaling language models to handle longer contexts introduces substantial memory challenges due to the growing cost of key-value (KV) caches. Motivated by the efficiency gains of hybrid models and the broad availability of pretrained large transformer backbones, we explore transitioning transformer models into hybrid architectures for a more efficient generation. In this work, we propose LightTransfer, a lightweight method that transforms models such as LLaMA into hybrid variants. Our approach identifies lazy layers -- those focusing on recent or initial tokens -- and replaces their full attention with streaming attention. This transformation can be performed without any training for long-context understanding tasks or with minimal fine-tuning for o1-like long reasoning generation tasks that require stronger reasoning capabilities. Experiments across diverse benchmarks and models (e.g., LLaMA, Mistral, QwQ-STILL) demonstrate that, even when half of the layers are identified as lazy, LightTransfer achieves up to 2.17times throughput improvement with minimal performance loss (<1.5% on LongBench) and achieves 53.3\% on math benchmark AIME24 of advanced o1-like long reasoning model QwQ-STILL.

Spoken dialog systems with cascaded ASR-LLM-TTS modules retain strong LLM intelligence, but VAD segmentation often forces half-duplex turns and brittle control. On the other hand, VAD-free end-to-end model support full-duplex interaction but is hard to maintain conversational intelligence. In this paper, we present DuplexCascade, a VAD-free cascaded streaming pipeline for full-duplex speech-to-speech dialogue. Our key idea is to convert conventional utterance-wise long turns into chunk-wise micro-turn interactions, enabling rapid bidirectional exchange while preserving the strengths of a capable text LLM. To reliably coordinate turn-taking and response timing, we introduce a set of conversational special control tokens that steer the LLM's behavior under streaming constraints. On Full-DuplexBench and VoiceBench, DuplexCascade delivers state-of-the-art full-duplex turn-taking and strong conversational intelligence among open-source speech-to-speech dialogue systems.

We present Kimi-Audio, an open-source audio foundation model that excels in audio understanding, generation, and conversation. We detail the practices in building Kimi-Audio, including model architecture, data curation, training recipe, inference deployment, and evaluation. Specifically, we leverage a 12.5Hz audio tokenizer, design a novel LLM-based architecture with continuous features as input and discrete tokens as output, and develop a chunk-wise streaming detokenizer based on flow matching. We curate a pre-training dataset that consists of more than 13 million hours of audio data covering a wide range of modalities including speech, sound, and music, and build a pipeline to construct high-quality and diverse post-training data. Initialized from a pre-trained LLM, Kimi-Audio is continual pre-trained on both audio and text data with several carefully designed tasks, and then fine-tuned to support a diverse of audio-related tasks. Extensive evaluation shows that Kimi-Audio achieves state-of-the-art performance on a range of audio benchmarks including speech recognition, audio understanding, audio question answering, and speech conversation. We release the codes, model checkpoints, as well as the evaluation toolkits in https://github.com/MoonshotAI/Kimi-Audio.

It is well known that LLM-based systems are data-hungry. Recent LLM-based TTS works typically employ complex data processing pipelines to obtain high-quality training data. These sophisticated pipelines require excellent models at each stage (e.g., speech denoising, speech enhancement, speaker diarization, and punctuation models), which themselves demand high-quality training data and are rarely open-sourced. Even with state-of-the-art models, issues persist, such as incomplete background noise removal and misalignment between punctuation and actual speech pauses. Moreover, the stringent filtering strategies often retain only 10-30\% of the original data, significantly impeding data scaling efforts. In this work, we leverage a noise-robust audio tokenizer (S3Tokenizer) to design a simplified yet effective TTS data processing pipeline that maintains data quality while substantially reducing data acquisition costs, achieving a data retention rate of over 50\%. Beyond data scaling challenges, LLM-based TTS systems also incur higher deployment costs compared to conventional approaches. Current systems typically use LLMs solely for text-to-token generation, while requiring separate models (e.g., flow matching models) for token-to-waveform generation, which cannot be directly executed by LLM inference engines, further complicating deployment. To address these challenges, we eliminate redundant modules in both LLM and flow components, replacing the flow model backbone with an LLM architecture. Building upon this simplified flow backbone, we propose a unified architecture for both streaming and non-streaming inference, significantly reducing deployment costs. Finally, we explore the feasibility of unifying TTS and ASR tasks using the same data for training, thanks to the simplified pipeline and the S3Tokenizer that reduces the quality requirements for TTS training data.

Large Language Models (LLMs) are primarily designed for batch processing. Existing methods for adapting LLMs to streaming rely either on expensive re-encoding or specialized architectures with limited scalability. This work identifies three key mismatches in adapting batch-oriented LLMs to streaming: (1) input-attention, (2) output-attention, and (3) position-ID mismatches. While it is commonly assumed that the latter two mismatches require frequent re-encoding, our analysis reveals that only the input-attention mismatch significantly impacts performance, indicating re-encoding outputs is largely unnecessary. To better understand this discrepancy with the common assumption, we provide the first comprehensive analysis of the impact of position encoding on LLMs in streaming, showing that preserving relative positions within source and target contexts is more critical than maintaining absolute order. Motivated by the above analysis, we introduce a group position encoding paradigm built on batch architectures to enhance consistency between streaming and batch modes. Extensive experiments on cross-lingual and cross-modal tasks demonstrate that our method outperforms existing approaches. Our method requires no architectural modifications, exhibits strong generalization in both streaming and batch modes. The code is available at repository https://github.com/EIT-NLP/StreamingLLM.

Recent Large Language Models have been enhanced with vision capabilities, enabling them to comprehend images, videos, and interleaved vision-language content. However, the learning methods of these large multimodal models typically treat videos as predetermined clips, making them less effective and efficient at handling streaming video inputs. In this paper, we propose a novel Learning-In-Video-Stream (LIVE) framework, which enables temporally aligned, long-context, and real-time conversation within a continuous video stream. Our LIVE framework comprises comprehensive approaches to achieve video streaming dialogue, encompassing: (1) a training objective designed to perform language modeling for continuous streaming inputs, (2) a data generation scheme that converts offline temporal annotations into a streaming dialogue format, and (3) an optimized inference pipeline to speed up the model responses in real-world video streams. With our LIVE framework, we built VideoLLM-online model upon Llama-2/Llama-3 and demonstrate its significant advantages in processing streaming videos. For instance, on average, our model can support streaming dialogue in a 5-minute video clip at over 10 FPS on an A100 GPU. Moreover, it also showcases state-of-the-art performance on public offline video benchmarks, such as recognition, captioning, and forecasting. The code, model, data, and demo have been made available at https://showlab.github.io/videollm-online.

Generative tasks, such as text generation and question answering, hold a crucial position in the realm of mobile applications. Due to their sensitivity to privacy concerns, there is a growing demand for their execution directly on mobile devices. Currently, the execution of these generative tasks heavily depends on Large Language Models (LLMs). Nevertheless, the limited memory capacity of these devices presents a formidable challenge to the scalability of such models. In our research, we introduce LLMCad, an innovative on-device inference engine specifically designed for efficient generative Natural Language Processing (NLP) tasks. The core idea behind LLMCad revolves around model collaboration: a compact LLM, residing in memory, takes charge of generating the most straightforward tokens, while a high-precision LLM steps in to validate these tokens and rectify any identified errors. LLMCad incorporates three novel techniques: (1) Instead of generating candidate tokens in a sequential manner, LLMCad employs the smaller LLM to construct a token tree, encompassing a wider range of plausible token pathways. Subsequently, the larger LLM can efficiently validate all of these pathways simultaneously. (2) It employs a self-adjusting fallback strategy, swiftly initiating the verification process whenever the smaller LLM generates an erroneous token. (3) To ensure a continuous flow of token generation, LLMCad speculatively generates tokens during the verification process by implementing a compute-IO pipeline. Through an extensive series of experiments, LLMCad showcases an impressive token generation speed, achieving rates up to 9.3x faster than existing inference engines.

Recent works have shown that large language model (LLM) agents are able to improve themselves from experience, which is an important ability for continuous enhancement post-deployment. However, existing benchmarks primarily evaluate their innate capabilities and do not assess their ability to improve over time. To address this gap, we introduce StreamBench, a pioneering benchmark designed to evaluate the continuous improvement of LLM agents over an input-feedback sequence. StreamBench simulates an online learning environment where LLMs receive a continuous flow of feedback stream and iteratively enhance their performance. In addition, we propose several simple yet effective baselines for improving LLMs on StreamBench, and provide a comprehensive analysis to identify critical components that contribute to successful streaming strategies. Our work serves as a stepping stone towards developing effective online learning strategies for LLMs, paving the way for more adaptive AI systems in streaming scenarios.

Providers of LLM-as-a-service have predominantly adopted a simple pricing model: users pay a fixed price per token. Consequently, one may think that the price two different users would pay for the same output string under the same input prompt is the same. In our work, we show that, surprisingly, this is not (always) true. We find empirical evidence that, particularly for non-english outputs, both proprietary and open-weights LLMs often generate the same (output) string with multiple different tokenizations, even under the same input prompt, and this in turn leads to arbitrary price variation. To address the problem of tokenization multiplicity, we introduce canonical generation, a type of constrained generation that restricts LLMs to only generate canonical tokenizations -- the unique tokenization in which each string is tokenized during the training process of an LLM. Further, we introduce an efficient sampling algorithm for canonical generation based on the Gumbel-Max trick. Experiments on a variety of natural language tasks demonstrate that our sampling algorithm for canonical generation is comparable to standard sampling in terms of performance and runtime, and it solves the problem of tokenization multiplicity.

Large language models (LLMs) are known for their exceptional performance across a range of natural language processing tasks, but their deployment comes at a high computational and financial cost. On the other hand, smaller language models (SLMs), which can be deployed on lower-cost edge devices, struggle to match the performance of their larger counterparts. This paper presents a novel hybrid inference approach that leverages the strengths of both model types while minimizing reliance on costly cloud-based LLMs. Unlike existing methods that route entire queries to either an SLM or a cloud LLM, our approach introduces a reward-based mechanism to dynamically determine the involvement of the cloud LLM during token generation. Specifically, each token predicted by the SLM is evaluated against a reward score, and only when this score falls below a certain threshold is the cloud LLM consulted for assistance in the next token prediction. This method not only reduces the traffic to the cloud LLM, thereby lowering costs, but also allows for flexible control over response quality depending on the reward score threshold. Experimental results demonstrate that our approach significantly reduces cloud LLM usage with minimal impact on overall response quality, offering a cost-effective solution for deploying high-performance language models

Tokenization is an understudied and often neglected component of modern LLMs. Most published works use a single tokenizer for all experiments, often borrowed from another model, without performing ablations or analysis to optimize tokenization. Moreover, the tokenizer is generally kept unchanged when fine-tuning a base model. In this paper, we show that the size, pre-tokenization regular expression, and training data of a tokenizer can significantly impact the model's generation speed, effective context size, memory usage, and downstream performance. We train specialized Byte-Pair Encoding code tokenizers, and conduct extensive ablations on the impact of tokenizer design on the performance of LLMs for code generation tasks such as HumanEval and MBPP, and provide recommendations for tokenizer hyper-parameters selection and switching the tokenizer in a pre-trained LLM. We perform our experiments on models trained from scratch and from pre-trained models, verifying their applicability to a wide range of use-cases. We find that when fine-tuning on more than 50 billion tokens, we can specialize the tokenizer of a pre-trained LLM to obtain large gains in generation speed and effective context size.

Video large language models (Video-LLMs) have demonstrated strong capabilities in video understanding tasks. However, their practical deployment is still hindered by the inefficiency introduced by processing massive amounts of visual tokens. Although recent approaches achieve extremely low token retention ratios while maintaining accuracy comparable to full-token baselines, most of them perform compression only at the late stage of prefilling, leaving the efficiency of the vision encoder unoptimized. In this paper, we first show that vision encoding contributes a large portion to the time-to-first-token (TTFT). Therefore, instead of compressing visual tokens only after the vision encoder, performing compression inside the encoder still leaves substantial room for exploration. Based on this insight, we propose EarlyTom, a training-free token compression framework that performs early-stage visual token compression inside the vision encoder, enabling significantly better TTFT reduction and higher throughput. In addition, we introduce a decoupled spatial token selection strategy that improves the overall compression effectiveness. EarlyTom reduces TTFT by up to 2.65x and FLOPs by up to 61% on a single NVIDIA A100 GPU for the LLaVA-OneVision-7B model, while maintaining accuracy comparable to the full-token baseline. These improvements substantially enhance the practicality of deploying Video-LLMs in real-world production scenarios.

Large language models (LLMs) propel the prosperity of interactive AI applications showcased by ChatGPT that demand timely response of inference services. However, LLM inference is computation intensive and memory intensive, and improper parameter configuration at LLM platforms may exacerbate the inference time. In this paper, we analyze the impact of LLM output token distribution on the inference queueing delay, where the max-token clipping and the batched inference are considered. By formulating an M/G/1 model, we observe that enforcing a maximum output token limit on a very small fraction of inference requests can significantly reduce the queueing delay, and our model facilitates the selection of the optimal limit. For the batch inference, we model the service process as a bulk queue in which the batch processing time is affected by the batch size and the maximum token size inside this batch jointly. The queueing delays of the batching of all buffered requests (dynamic batching), the batching of constant number of requests (fixed batching), and the batching without intra-batch waiting (elastic batching) are derived. Experimental results show that our mathematical models coincide with the event-driven simulations well.

The complexity of large language model (LLM) serving workloads has substantially increased due to the integration with external tool invocations, such as ChatGPT plugins. In this paper, we identify a new opportunity for efficient LLM serving for requests that trigger tools: tool partial execution alongside LLM decoding. To this end, we design Conveyor, an efficient LLM serving system optimized for handling requests involving external tools. We introduce a novel interface for tool developers to expose partial execution opportunities to the LLM serving system and a request scheduler that facilitates partial tool execution. Our results demonstrate that tool partial execution can improve request completion latency by up to 38.8%.

Large Language Models (LLMs) have showcased impressive capabilities in text comprehension and generation, prompting research efforts towards video LLMs to facilitate human-AI interaction at the video level. However, how to effectively encode and understand videos in video-based dialogue systems remains to be solved. In this paper, we investigate a straightforward yet unexplored question: Can we feed all spatial-temporal tokens into the LLM, thus delegating the task of video sequence modeling to the LLMs? Surprisingly, this simple approach yields significant improvements in video understanding. Based upon this, we propose ST-LLM, an effective video-LLM baseline with Spatial-Temporal sequence modeling inside LLM. Furthermore, to address the overhead and stability issues introduced by uncompressed video tokens within LLMs, we develop a dynamic masking strategy with tailor-made training objectives. For particularly long videos, we have also designed a global-local input module to balance efficiency and effectiveness. Consequently, we harness LLM for proficient spatial-temporal modeling, while upholding efficiency and stability. Extensive experimental results attest to the effectiveness of our method. Through a more concise model and training pipeline, ST-LLM establishes a new state-of-the-art result on VideoChatGPT-Bench and MVBench. Codes have been available at https://github.com/TencentARC/ST-LLM.

We present StreamBridge, a simple yet effective framework that seamlessly transforms offline Video-LLMs into streaming-capable models. It addresses two fundamental challenges in adapting existing models into online scenarios: (1) limited capability for multi-turn real-time understanding, and (2) lack of proactive response mechanisms. Specifically, StreamBridge incorporates (1) a memory buffer combined with a round-decayed compression strategy, supporting long-context multi-turn interactions, and (2) a decoupled, lightweight activation model that can be effortlessly integrated into existing Video-LLMs, enabling continuous proactive responses. To further support StreamBridge, we construct Stream-IT, a large-scale dataset tailored for streaming video understanding, featuring interleaved video-text sequences and diverse instruction formats. Extensive experiments show that StreamBridge significantly improves the streaming understanding capabilities of offline Video-LLMs across various tasks, outperforming even proprietary models such as GPT-4o and Gemini 1.5 Pro. Simultaneously, it achieves competitive or superior performance on standard video understanding benchmarks.

Large Language Models (LLMs) have exhibited exceptional performance across a spectrum of natural language processing tasks. However, their substantial sizes pose considerable challenges, particularly in computational demands and inference speed, due to their quadratic complexity. In this work, we have identified a key pattern: certain seemingly meaningless special tokens (i.e., separators) contribute disproportionately to attention scores compared to semantically meaningful tokens. This observation suggests that information of the segments between these separator tokens can be effectively condensed into the separator tokens themselves without significant information loss. Guided by this insight, we introduce SepLLM, a plug-and-play framework that accelerates inference by compressing these segments and eliminating redundant tokens. Additionally, we implement efficient kernels for training acceleration. Experimental results across training-free, training-from-scratch, and post-training settings demonstrate SepLLM's effectiveness. Notably, using the Llama-3-8B backbone, SepLLM achieves over 50% reduction in KV cache on the GSM8K-CoT benchmark while maintaining comparable performance. Furthermore, in streaming settings, SepLLM effectively processes sequences of up to 4 million tokens or more while maintaining consistent language modeling capabilities.

The rapid proliferation of LLM agent frameworks has forced developers to choose between vendor lock-in through provider-specific SDKs and complex multi-package ecosystems that obscure control flow and hinder reproducibility. Integrating tool calling across multiple LLM providers remains a core engineering challenge due to fragmented APIs, incompatible message formats, and inconsistent streaming and tool-calling behavior, making it difficult to build portable, reliable agent systems. We introduce Orchestral, a lightweight Python framework that provides a unified, type-safe interface for building LLM agents across major providers while preserving the simplicity required for scientific computing and production deployment. Orchestral defines a single universal representation for messages, tools, and LLM usage that operates seamlessly across providers, eliminating manual format translation and reducing framework-induced complexity. Automatic tool schema generation from Python type hints removes the need for handwritten descriptors while maintaining type safety across provider boundaries. A synchronous execution model with streaming support enables deterministic behavior, straightforward debugging, and real-time interaction without introducing server dependencies. The framework's modular architecture cleanly separates provider integration, tool execution, conversation orchestration, and user-facing interfaces, enabling extensibility without architectural entanglement. Orchestral supports advanced agent capabilities found in larger frameworks, including rich tool calling, context compaction, workspace sandboxing, user approval workflows, sub-agents, memory management, and MCP integration.

The rise of large language models (LLMs) has enabled LLM-based applications (a.k.a. AI agents or co-pilots), a new software paradigm that combines the strength of LLM and conventional software. Diverse LLM applications from different tenants could design complex workflows using multiple LLM requests to accomplish one task. However, they have to use the over-simplified request-level API provided by today's public LLM services, losing essential application-level information. Public LLM services have to blindly optimize individual LLM requests, leading to sub-optimal end-to-end performance of LLM applications. This paper introduces Parrot, an LLM service system that focuses on the end-to-end experience of LLM-based applications. Parrot proposes Semantic Variable, a unified abstraction to expose application-level knowledge to public LLM services. A Semantic Variable annotates an input/output variable in the prompt of a request, and creates the data pipeline when connecting multiple LLM requests, providing a natural way to program LLM applications. Exposing Semantic Variables to the public LLM service allows it to perform conventional data flow analysis to uncover the correlation across multiple LLM requests. This correlation opens a brand-new optimization space for the end-to-end performance of LLM-based applications. Extensive evaluations demonstrate that Parrot can achieve up to an order-of-magnitude improvement for popular and practical use cases of LLM applications.

Large language model (LLM) applications are evolving beyond simple chatbots into dynamic, general-purpose agentic programs, which scale LLM calls and output tokens to help AI agents reason, explore, and solve complex tasks. However, existing LLM serving systems ignore dependencies between programs and calls, missing significant opportunities for optimization. Our analysis reveals that programs submitted to LLM serving engines experience long cumulative wait times, primarily due to head-of-line blocking at both the individual LLM request and the program. To address this, we introduce Autellix, an LLM serving system that treats programs as first-class citizens to minimize their end-to-end latencies. Autellix intercepts LLM calls submitted by programs, enriching schedulers with program-level context. We propose two scheduling algorithms-for single-threaded and distributed programs-that preempt and prioritize LLM calls based on their programs' previously completed calls. Our evaluation demonstrates that across diverse LLMs and agentic workloads, Autellix improves throughput of programs by 4-15x at the same latency compared to state-of-the-art systems, such as vLLM.

LLM inference is still evaluated mainly as a model or software problem: accuracy, latency, throughput, and hardware utilization. This is incomplete. At deployment scale, the relevant output is a quality-conditioned token produced under joint constraints from effective compute, delivered data-center power, cooling capacity, PUE, and utilization. We argue that the ML community should treat inference as energy-to-token production. We formalize this view with a dimensionally consistent Token Production Function in which token rate is bounded by both compute-per-token and energy-per-token ceilings. Listed API prices vary by over an order of magnitude across providers, but we use price dispersion only as directional motivation, not as causal evidence of marginal cost. The core physical question is instead: under fixed quality and service targets, when does the binding constraint move from theoretical peak compute toward delivered power, cooling, and operational efficiency? Under this framing, system optimizations -- latent KV-cache compression, sparse or heavily compressed attention, quantization, routing, and difficulty-adaptive reasoning -- are not merely local engineering tricks. They are energy-to-token levers because they reduce FLOPs/token, joules/token, memory traffic, or utilization losses under fixed (q^{*},s^{*}). We therefore call for inference papers and benchmarks to report Joules/token, active binding constraint, PUE-adjusted delivered power, and utilization-adjusted token output alongside accuracy and latency.

Large Language Models (LLMs) are increasingly applied to data-intensive workflows, from database querying to developer observability. Yet the effectiveness of these systems is constrained by the volume, verbosity, and noise of real-world text-rich data such as logs, telemetry, and monitoring streams. Feeding such data directly into LLMs is costly, environmentally unsustainable, and often misaligned with task objectives. Parallel efforts in LLM efficiency have focused on model- or architecture-level optimizations, but the challenge of reducing upstream input verbosity remains underexplored. In this paper, we argue for treating the token budget of an LLM as an attention budget and elevating task-aware text reduction as a first-class design principle for language -- data systems. We position input-side reduction not as compression, but as attention allocation: prioritizing information most relevant to downstream tasks. We outline open research challenges for building benchmarks, designing adaptive reduction pipelines, and integrating token-budget--aware preprocessing into database and retrieval systems. Our vision is to channel scarce attention resources toward meaningful signals in noisy, data-intensive workflows, enabling scalable, accurate, and sustainable LLM--data integration.

We present OnPrem.LLM, a Python-based toolkit for applying large language models (LLMs) to sensitive, non-public data in offline or restricted environments. The system is designed for privacy-preserving use cases and provides prebuilt pipelines for document processing and storage, retrieval-augmented generation (RAG), information extraction, summarization, classification, and prompt/output processing with minimal configuration. OnPrem.LLM supports multiple LLM backends -- including llama.cpp, Ollama, vLLM, and Hugging Face Transformers -- with quantized model support, GPU acceleration, and seamless backend switching. Although designed for fully local execution, OnPrem.LLM also supports integration with a wide range of cloud LLM providers when permitted, enabling hybrid deployments that balance performance with data control. A no-code web interface extends accessibility to non-technical users.

The rapid advancement of Large Language Models (LLMs) has significantly impacted human-computer interaction, epitomized by the release of GPT-4o, which introduced comprehensive multi-modality capabilities. In this paper, we first explored the deployment strategies, economic considerations, and sustainability challenges associated with the state-of-the-art LLMs. More specifically, we discussed the deployment debate between Retrieval-Augmented Generation (RAG) and fine-tuning, highlighting their respective advantages and limitations. After that, we quantitatively analyzed the requirement of xPUs in training and inference. Additionally, for the tokenomics of LLM services, we examined the balance between performance and cost from the quality of experience (QoE)'s perspective of end users. Lastly, we envisioned the future hybrid architecture of LLM processing and its corresponding sustainability concerns, particularly in the environmental carbon footprint impact. Through these discussions, we provided a comprehensive overview of the operational and strategic considerations essential for the responsible development and deployment of LLMs.

The wide adoption of LLMs has led to their use in great variety of applications and scenarios, such as chatbot assistants and data annotation, creating the need for the models to satisfy certain budget and hardware constraints. This has led to the trend of LLMs being released in batches consisting of similar models of various sizes for the family of models to adhere to as wide of a range of constraints as possible. In this paper, we validate distillation and quantization as a cost-effective way to expand model families to new sizes and hardware formats. Based on the open-recipe Apertus 8B LLM, we produce Apertus-v1.1 - a distilled family of models with up to 4B parameters trained on 1.7T permissive license tokens. We demonstrate cost-efficiency and strong accuracy performance of our approach for covering large ranges of hardware and systems requirements.

Speech Large Language Models (Speech LLMs) lack a principled mechanism for streaming inference: their label-synchronous generation has no acoustic-frame alignment, making real-time decoding and end-of-utterance detection difficult. We propose TRADE TRansducer-Augmented DEcoder, which augments a multimodal LLM with a transducer branch that shares the audio encoder and uses the LLM's hidden states directly as the prediction network -- coupling frame-synchronous acoustic alignment with the LLM's linguistic reasoning. Three design choices make the system accurate, streamable, and long-form capable: (1)Tightly coupled dual vocabularies -- a compact transducer vocabulary derived from the LLM vocabulary, enabling zero-cost score fusion; (2)Chunk-synchronized streaming training with gradient stopping, eliminating the train-inference mismatch at offline-equivalent memory cost; and (3)Localized Decoder Audio Attention (LDAA), a causal sliding window that caps KV-cache memory independently of utterance length. A single TRADE checkpoint supports offline and streaming decoding across a continuous range of latency operating points. TRADE achieves 6.71% average WER on the Open ASR Leaderboard, while the streaming recognition with 960ms chunk size reaches 8.40% from the same checkpoint. On long-form speech, it obtains 3.64% WER on TED-LIUM and 10.88% on Earnings-22 without external segmentation. TRADE provides sentence-end punctuation timestamps that, when combined with acoustic voice activity detection (VAD), improve end-of-utterance detection by +0.03 F_1 over acoustic VAD alone.

Transformer based Large Language Models (LLMs) have been widely used in many fields, and the efficiency of LLM inference becomes hot topic in real applications. However, LLMs are usually complicatedly designed in model structure with massive operations and perform inference in the auto-regressive mode, making it a challenging task to design a system with high efficiency. In this paper, we propose an efficient LLM inference solution with low latency and high throughput. Firstly, we simplify the LLM decoder layer by fusing data movement and element-wise operations to reduce the memory access frequency and lower system latency. We also propose a segment KV cache policy to keep key/value of the request and response tokens in separate physical memory for effective device memory management, helping enlarge the runtime batch size and improve system throughput. A customized Scaled-Dot-Product-Attention kernel is designed to match our fusion policy based on the segment KV cache solution. We implement our LLM inference solution on Intel GPU and publish it publicly. Compared with the standard HuggingFace implementation, the proposed solution achieves up to 7x lower token latency and 27x higher throughput for some popular LLMs on Intel GPU.

Video Large Language Models (VideoLLMs) have achieved strong performance on many video understanding tasks, but most existing systems remain offline and are not well-suited for live video streams that require continuous observation and timely response. Recent streaming VideoLLMs have made progress, yet current approaches often rely on decoupled trigger-response pipelines or are limited to captioning-style narration, reducing their effectiveness for open-ended question answering and long-horizon interaction. We propose AURA (Always-On Understanding and Real-Time Assistance), an end-to-end streaming visual interaction framework that enables a unified VideoLLM to continuously process video streams and support both real-time question answering and proactive responses. AURA integrates context management, data construction, training objectives, and deployment optimization for stable long-horizon streaming interaction. It achieves state-of-the-art performance on streaming benchmarks and supports a real-time demo system with ASR and TTS running at 2 FPS on two 80G accelerators. We release the AURA model together with a real-time inference framework to facilitate future research.

We study how to watermark LLM outputs, i.e. embedding algorithmically detectable signals into LLM-generated text to track misuse. Unlike the current mainstream methods that work with a fixed LLM, we expand the watermark design space by including the LLM tuning stage in the watermark pipeline. While prior works focus on token-level watermark that embeds signals into the output, we design a model-level watermark that embeds signals into the LLM weights, and such signals can be detected by a paired detector. We propose a co-training framework based on reinforcement learning that iteratively (1) trains a detector to detect the generated watermarked text and (2) tunes the LLM to generate text easily detectable by the detector while keeping its normal utility. We empirically show that our watermarks are more accurate, robust, and adaptable (to new attacks). It also allows watermarked model open-sourcing. In addition, if used together with alignment, the extra overhead introduced is low - only training an extra reward model (i.e. our detector). We hope our work can bring more effort into studying a broader watermark design that is not limited to working with a fixed LLM. We open-source the code: https://github.com/xiaojunxu/learning-to-watermark-llm .

LLM-based agents are increasingly capable of complex task execution, yet current agentic systems remain constrained by text-centric paradigms. Traditional approaches rely on procedural JSON-based function calling, which often struggles with long-horizon tasks due to fragile multi-turn dependencies and context drift. In this paper, we present CaveAgent, a framework that transforms the paradigm from "LLM-as-Text-Generator" to "LLM-as-Runtime-Operator." We introduce a Dual-stream Context Architecture that decouples state management into a lightweight semantic stream for reasoning and a persistent, deterministic Python Runtime stream for execution. In addition to leveraging code generation to efficiently resolve interdependent sub-tasks (e.g., loops, conditionals) in a single step, we introduce Stateful Runtime Management in CaveAgent. Distinct from existing code-based approaches that remain text-bound and lack the support for external object injection and retrieval, CaveAgent injects, manipulates, and retrieves complex Python objects (e.g., DataFrames, database connections) that persist across turns. This persistence mechanism acts as a high-fidelity external memory to eliminate context drift, avoid catastrophic forgetting, while ensuring that processed data flows losslessly to downstream applications. Comprehensive evaluations on Tau^2-bench, BFCL and various case studies across representative SOTA LLMs demonstrate CaveAgent's superiority. Specifically, our framework achieves a 10.5\% success rate improvement on retail tasks and reduces total token consumption by 28.4\% in multi-turn scenarios. On data-intensive tasks, direct variable storage and retrieval reduces token consumption by 59\%, allowing CaveAgent to handle large-scale data that causes context overflow failures in both JSON-based and Code-based agents.

The computational and energy costs of Large Language Models (LLMs) have increased exponentially driven by the growing model sizes and the massive adoption of LLMs by hundreds of millions of users. The unit cost of an LLM is the computation of a token. Therefore, the tokenizer plays an important role in the efficiency of a model, and they are carefully optimized to minimize the number of tokens for the text in their training corpus. One of the most popular applications of LLMs are chatbots that interact with users. A key observation is that, for those chatbots, what is important is the performance of the tokenizer in the user text input and the chatbot responses. Those are most likely different from the text in the training corpus. So, a question that immediately arises is whether there is a potential benefit in optimizing tokenizers for chatbot conversations. In this paper, this idea is explored for different tokenizers by using a publicly available corpus of chatbot conversations to redesign their vocabularies and evaluate their performance in this domain. The results show that conversation-optimized tokenizers consistently reduce the number of tokens in chatbot dialogues, which can lead to meaningful energy savings, in the range of 5% to 10% while having minimal or even slightly positive impact on tokenization efficiency for the original training corpus.

Multi-agent LLM applications organize execution in synchronized rounds where a central scheduler gathers outputs from all agents and redistributes the combined context. This All-Gather communication pattern creates massive KV Cache redundancy, because every agent's prompt contains the same shared output blocks, yet existing reuse methods fail to exploit it efficiently. We present TokenDance, a system that scales the number of concurrent agents by exploiting the All-Gather pattern for collective KV Cache sharing. TokenDance's KV Collector performs KV Cache reuse over the full round in one collective step, so the cost of reusing a shared block is paid once regardless of agent count. Its Diff-Aware Storage encodes sibling caches as block-sparse diffs against a single master copy, achieving 11-17x compression on representative workloads. Evaluation on GenerativeAgents and AgentSociety shows that TokenDance supports up to 2.7x more concurrent agents than vLLM with prefix caching under SLO requirement, reduces per-agent KV Cache storage by up to 17.5x, and achieves up to 1.9x prefill speedup over per-request position-independent caching.

Large Language Models (LLMs) have demonstrated remarkable capabilities in handling long context inputs, but this comes at the cost of increased computational resources and latency. Our research introduces a novel approach for the long context bottleneck to accelerate LLM inference and reduce GPU memory consumption. Our research demonstrates that LLMs can identify relevant tokens in the early layers before generating answers to a query. Leveraging this insight, we propose an algorithm that uses early layers of an LLM as filters to select and compress input tokens, significantly reducing the context length for subsequent processing. Our method, GemFilter, demonstrates substantial improvements in both speed and memory efficiency compared to existing techniques, such as standard attention and SnapKV/H2O. Notably, it achieves a 2.4times speedup and 30\% reduction in GPU memory usage compared to SOTA methods. Evaluation on the Needle in a Haystack task shows that GemFilter significantly outperforms standard attention, SnapKV and demonstrates comparable performance on the LongBench challenge. GemFilter is simple, training-free, and broadly applicable across different LLMs. Crucially, it provides interpretability by allowing humans to inspect the selected input sequence. These findings not only offer practical benefits for LLM deployment, but also enhance our understanding of LLM internal mechanisms, paving the way for further optimizations in LLM design and inference. Our code is available at https://github.com/SalesforceAIResearch/GemFilter.

In this paper, we establish the novel insight that an off-the-shelf LLM can function as an excellent token compressor and decompressor. To demonstrate, we design a self-expressive autoencoding learning framework fine-tunes a pretrained LLM to translate long texts into a compact internal language of discrete, variable-length latent codes, termed Z-tokens, and to reconstruct the original text exactly from them. The resulting representation is content-adaptive: semantically dense segments receive more Z-tokens, while redundant or predictable regions are aggressively compressed, via lightweight LoRA-based adapter heads. Empirically, our method achieves up to 18 times token reduction on Wikipedia, CNN/DailyMail, HotpotQA, and Qulac-style long-query datasets, while preserving reconstruction fidelity and downstream performance. This simple yet effective design supports applications including prompt compression and autoregressive generation directly in the Z-token space, offering a potential pathway toward token-efficient long-context reasoning.

The recent success of LLMs has been predominantly driven by curating the training dataset composition, scaling of model architectures and dataset sizes and advancements in pretraining objectives, leaving tokenizer influence as a blind spot. Shedding light on this underexplored area, we conduct a comprehensive study on the influence of tokenizer choice on LLM downstream performance by training 24 mono- and multilingual LLMs at a 2.6B parameter scale, ablating different tokenizer algorithms and parameterizations. Our studies highlight that the tokenizer choice can significantly impact the model's downstream performance, training and inference costs. In particular, we find that the common tokenizer evaluation metrics fertility and parity are not always predictive of model downstream performance, rendering these metrics a questionable proxy for the model's downstream performance. Furthermore, we show that multilingual tokenizers trained on the five most frequent European languages require vocabulary size increases of factor three in comparison to English. While English-only tokenizers have been applied to the training of multi-lingual LLMs, we find that this approach results in a severe downstream performance degradation and additional training costs of up to 68%, due to an inefficient tokenization vocabulary.

We introduce ProVideLLM, an end-to-end framework for real-time procedural video understanding. ProVideLLM integrates a multimodal cache configured to store two types of tokens - verbalized text tokens, which provide compressed textual summaries of long-term observations, and visual tokens, encoded with DETR-QFormer to capture fine-grained details from short-term observations. This design reduces token count by 22x over existing methods in representing one hour of long-term observations while effectively encoding fine-granularity of the present. By interleaving these tokens in our multimodal cache, ProVideLLM ensures sub-linear scaling of memory and compute with video length, enabling per-frame streaming inference at 10 FPS and streaming dialogue at 25 FPS, with a minimal 2GB GPU memory footprint. ProVideLLM also sets new state-of-the-art results on six procedural tasks across four datasets.

While Large Language Models (LLMs) are the dominant models for generative tasks in language, they do not perform as well as diffusion models on image and video generation. To effectively use LLMs for visual generation, one crucial component is the visual tokenizer that maps pixel-space inputs to discrete tokens appropriate for LLM learning. In this paper, we introduce MAGVIT-v2, a video tokenizer designed to generate concise and expressive tokens for both videos and images using a common token vocabulary. Equipped with this new tokenizer, we show that LLMs outperform diffusion models on standard image and video generation benchmarks including ImageNet and Kinetics. In addition, we demonstrate that our tokenizer surpasses the previously top-performing video tokenizer on two more tasks: (1) video compression comparable to the next-generation video codec (VCC) according to human evaluations, and (2) learning effective representations for action recognition tasks.

Current LLM-based coding agents follow a serial execution paradigm: the model first generates the complete code, then invokes an interpreter to execute it. This sequential workflow leaves the executor idle during generation and the generator idle during execution, resulting in unnecessary end-to-end latency. We observe that, unlike human developers, LLMs produce code tokens sequentially without revision, making it possible to execute code as it is being generated. We formalize this parallel execution paradigm, modeling it as a three-stage pipeline of generation, detection, and execution, and derive closed-form latency bounds that characterize its speedup potential and operating regimes. We then present Eager, a concrete implementation featuring AST-based chunking, dynamic batching with gated execution, and early error interruption. We evaluate Eager across four benchmarks, seven LLMs, and three execution environments. Results show that Eager reduces the non-overlapped execution latency by up to 99.9% and the end-to-end latency by up to 55% across seven LLMs and four benchmarks.

Video large language models (video LLMs) excel at video comprehension but face significant computational inefficiency due to redundant video tokens. Existing token pruning methods offer solutions. However, approaches operating within the LLM (inner-LLM pruning), such as FastV, incur intrinsic computational overhead in shallow layers. In contrast, methods performing token pruning before the LLM (outer-LLM pruning) primarily address spatial redundancy within individual frames or limited temporal windows, neglecting the crucial global temporal dynamics and correlations across longer video sequences. This leads to sub-optimal spatio-temporal reduction and does not leverage video compressibility fully. Crucially, the synergistic potential and mutual influence of combining these strategies remain unexplored. To further reduce redundancy, we introduce HoliTom, a novel training-free holistic token merging framework. HoliTom employs outer-LLM pruning through global redundancy-aware temporal segmentation, followed by spatial-temporal merging to reduce visual tokens by over 90%, significantly alleviating the LLM's computational burden. Complementing this, we introduce a robust inner-LLM token similarity-based merging approach, designed for superior performance and compatibility with outer-LLM pruning. Evaluations demonstrate our method's promising efficiency-performance trade-off on LLaVA-OneVision-7B, reducing computational costs to 6.9% of FLOPs while maintaining 99.1% of the original performance. Furthermore, we achieve a 2.28x reduction in Time-To-First-Token (TTFT) and a 1.32x acceleration in decoding throughput, highlighting the practical benefits of our integrated pruning approach for efficient video LLMs inference.

Diffusion Large Language Models (dLLMs) offer a compelling paradigm for natural language generation, leveraging parallel decoding and bidirectional attention to achieve superior global coherence compared to autoregressive models. While recent works have accelerated inference via KV cache reuse or heuristic decoding, they overlook the intrinsic inefficiencies within the block-wise diffusion process. Specifically, they suffer from spatial redundancy by modeling informative-sparse suffix regions uniformly and temporal inefficiency by applying fixed denoising schedules across all the decoding process. To address this, we propose Streaming-dLLM, a training-free framework that streamlines inference across both spatial and temporal dimensions. Spatially, we introduce attenuation guided suffix modeling to approximate the full context by pruning redundant mask tokens. Temporally, we employ a dynamic confidence aware strategy with an early exit mechanism, allowing the model to skip unnecessary iterations for converged tokens. Extensive experiments show that Streaming-dLLM achieves up to 68.2X speedup while maintaining generation quality, highlighting its effectiveness in diffusion decoding. The code is available at https://github.com/xiaoshideta/Streaming-dLLM.

Existing Multimodal Large Language Models (MLLMs) process a large number of visual tokens, leading to significant computational costs and inefficiency. Previous works generally assume that all visual tokens are necessary in the shallow layers of LLMs, and therefore token compression typically occurs in intermediate layers. In contrast, our study reveals an interesting insight: with proper selection, token compression is feasible at the input stage of LLM with negligible performance loss. Specifically, we reveal that explainability methods can effectively evaluate the importance of each visual token with respect to the given instruction, which can well guide the token compression. Furthermore, we propose to learn a mapping from the attention map of the first LLM layer to the explanation results, thereby avoiding the need for a full inference pass and facilitating practical deployment. Interestingly, this mapping can be learned using a simple and lightweight convolutional network, whose training is efficient and independent of MLLMs. Extensive experiments on 10 image and video benchmarks across three leading MLLMs (Qwen2-VL, LLaVA-OneVision, and VILA1.5) demonstrate the effectiveness of our approach, e.g., pruning 50% visual tokens while retaining more than 96% of the original performance across all benchmarks for all these three MLLMs. It also exhibits strong generalization, even when the number of tokens in inference far exceeds that used in training.

Recent innovations in generative large language models (LLMs) have made their applications and use-cases ubiquitous. This has led to large-scale deployments of these models, using complex, expensive, and power-hungry AI accelerators, most commonly GPUs. These developments make LLM inference efficiency an important challenge. Based on our extensive characterization, we find that there are two main phases during an LLM inference request: a compute-intensive prompt computation, and a memory-intensive token generation, each with distinct latency, throughput, memory, and power characteristics. Despite state-of-the-art batching and scheduling, the token generation phase underutilizes compute resources. Specifically, unlike compute-intensive prompt computation phases, token generation phases do not require the compute capability of the latest GPUs, and can be run with lower power and cost. With Splitwise, we propose splitting the two phases of a LLM inference request on to separate machines. This allows us to use hardware that is well-suited for each phase, and provision resources independently per phase. However, splitting an inference request across machines requires state transfer from the machine running prompt computation over to the machine generating tokens. We implement and optimize this state transfer using the fast back-plane interconnects available in today's GPU clusters. We use the Splitwise technique to design LLM inference clusters using the same or different types of machines for the prompt computation and token generation phases. Our clusters are optimized for three key objectives: throughput, cost, and power. In particular, we show that we can achieve 1.4x higher throughput at 20% lower cost than current designs. Alternatively, we can achieve 2.35x more throughput with the same cost and power budgets.

On-device large language models (LLMs) are catalyzing novel mobile applications such as UI task automation and personalized email auto-reply, without giving away users' private data. However, on-device LLMs still suffer from unacceptably long inference latency, especially the time to first token (prefill stage) due to the need of long context for accurate, personalized content generation, as well as the lack of parallel computing capacity of mobile CPU/GPU. To enable practical on-device LLM, we present mllm-NPU, the first-of-its-kind LLM inference system that efficiently leverages on-device Neural Processing Unit (NPU) offloading. Essentially, mllm-NPU is an algorithm-system co-design that tackles a few semantic gaps between the LLM architecture and contemporary NPU design. Specifically, it re-constructs the prompt and model in three levels: (1) At prompt level, it divides variable-length prompts into multiple fixed-sized chunks while maintaining data dependencies; (2) At tensor level, it identifies and extracts significant outliers to run on the CPU/GPU in parallel with minimal overhead; (3) At block level, it schedules Transformer blocks in an out-of-order manner to the CPU/GPU and NPU based on their hardware affinity and sensitivity to accuracy. Compared to competitive baselines, mllm-NPU achieves 22.4x faster prefill speed and 30.7x energy savings on average, and up to 32.8x speedup in an end-to-end real-world application. For the first time, mllm-NPU achieves more than 1,000 tokens/sec prefilling for a billion-sized model (Qwen1.5-1.8B), paving the way towards practical on-device LLM.

Video large language models (VideoLLM) excel at video understanding, but face efficiency challenges due to the quadratic complexity of abundant visual tokens. Our systematic analysis of token compression methods for VideoLLMs reveals two critical issues: (i) overlooking distinctive visual signals across frames, leading to information loss; (ii) suffering from implementation constraints, causing incompatibility with modern architectures or efficient operators. To address these challenges, we distill three design principles for VideoLLM token compression and propose a plug-and-play inference acceleration framework "Video Compression Commander" (VidCom2). By quantifying each frame's uniqueness, VidCom2 adaptively adjusts compression intensity across frames, effectively preserving essential information while reducing redundancy in video sequences. Extensive experiments across various VideoLLMs and benchmarks demonstrate the superior performance and efficiency of our VidCom2. With only 25% visual tokens, VidCom2 achieves 99.6% of the original performance on LLaVA-OV while reducing 70.8% of the LLM generation latency. Notably, our Frame Compression Adjustment strategy is compatible with other token compression methods to further improve their performance. Our code is available at https://github.com/xuyang-liu16/VidCom2.

In an era where micro-videos dominate platforms like TikTok and YouTube, AI-generated content is nearing cinematic quality. The next frontier is using large language models (LLMs) to autonomously create viral micro-videos, a largely untapped potential that could shape the future of AI-driven content creation. To address this gap, this paper presents the first exploration of LLM-assisted popular micro-video generation (LLMPopcorn). We selected popcorn as the icon for this paper because it symbolizes leisure and entertainment, aligning with this study on leveraging LLMs as assistants for generating popular micro-videos that are often consumed during leisure time. Specifically, we empirically study the following research questions: (i) How can LLMs be effectively utilized to assist popular micro-video generation? (ii) To what extent can prompt-based enhancements optimize the LLM-generated content for higher popularity? (iii) How well do various LLMs and video generators perform in the popular micro-video generation task? Exploring these questions, we show that advanced LLMs like DeepSeek-V3 can generate micro-videos with popularity rivaling human content. Prompt enhancement further boosts results, while benchmarking highlights DeepSeek-V3 and R1 for LLMs, and LTX-Video and HunyuanVideo for video generation. This work advances AI-assisted micro-video creation and opens new research directions. The code is publicly available at https://github.com/GAIR-Lab/LLMPopcorn.

Large Language Models (LLMs) have achieved remarkable performance and received significant research interest. The enormous computational demands, however, hinder the local deployment on devices with limited resources. The current prevalent LLM inference paradigms require users to send queries to the service providers for processing, which raises critical privacy concerns. Existing approaches propose to allow the users to obfuscate the token embeddings before transmission and utilize local models for denoising. Nonetheless, transmitting the token embeddings and deploying local models may result in excessive communication and computation overhead, preventing practical implementation. In this work, we propose DEL, a framework for Differentially private and communication Efficient LLM split inference. More specifically, an embedding projection module and a differentially private stochastic quantization mechanism are proposed to reduce the communication overhead in a privacy-preserving manner. To eliminate the need for local models, we adapt soft prompt at the server side to compensate for the utility degradation caused by privacy. To the best of our knowledge, this is the first work that utilizes soft prompt to improve the trade-off between privacy and utility in LLM inference, and extensive experiments on text generation and natural language understanding benchmarks demonstrate the effectiveness of the proposed method.

Large language models (LLMs) have been widely employed across various application domains, yet their black-box nature poses significant challenges to understanding how these models process input data internally to make predictions. In this paper, we introduce a precise and quantitative law that governs the learning of contextualized token embeddings through intermediate layers in pre-trained LLMs for next-token prediction. Our findings reveal that each layer contributes equally to enhancing prediction accuracy, from the lowest to the highest layer -- a universal phenomenon observed across a diverse array of open-source LLMs, built on architectures such as Transformer, RWKV, and Mamba. We demonstrate that this law offers new perspectives and insights to inform and guide practices in LLM development and applications, including model scaling, pre-training tasks, and information flow. Overall, our law enables more fine-grained approaches to the design, training, and interpretation of LLMs through scrutinizing their internal data processing mechanisms.

As large language models gain widespread adoption, running them efficiently becomes crucial. Recent works on LLM inference use speculative decoding to achieve extreme speedups. However, most of these works implicitly design their algorithms for high-end datacenter hardware. In this work, we ask the opposite question: how fast can we run LLMs on consumer machines? Consumer GPUs can no longer fit the largest available models (50B+ parameters) and must offload them to RAM or SSD. When running with offloaded parameters, the inference engine can process batches of hundreds or thousands of tokens at the same time as just one token, making it a natural fit for speculative decoding. We propose SpecExec (Speculative Execution), a simple parallel decoding method that can generate up to 20 tokens per target model iteration for popular LLM families. It utilizes the high spikiness of the token probabilities distribution in modern LLMs and a high degree of alignment between model output probabilities. SpecExec takes the most probable tokens continuation from the draft model to build a "cache" tree for the target model, which then gets validated in a single pass. Using SpecExec, we demonstrate inference of 50B+ parameter LLMs on consumer GPUs with RAM offloading at 4-6 tokens per second with 4-bit quantization or 2-3 tokens per second with 16-bit weights.

Large Language Model (LLM) routing is a pivotal technique for navigating a diverse landscape of LLMs, aiming to select the best-performing LLMs tailored to the domains of user queries, while managing computational resources. However, current routing approaches often face limitations in scalability when dealing with a large pool of specialized LLMs, or in their adaptability to extending model scope and evolving capability domains. To overcome those challenges, we propose InferenceDynamics, a flexible and scalable multi-dimensional routing framework by modeling the capability and knowledge of models. We operate it on our comprehensive dataset RouteMix, and demonstrate its effectiveness and generalizability in group-level routing using modern benchmarks including MMLU-Pro, GPQA, BigGenBench, and LiveBench, showcasing its ability to identify and leverage top-performing models for given tasks, leading to superior outcomes with efficient resource utilization. The broader adoption of Inference Dynamics can empower users to harness the full specialized potential of the LLM ecosystem, and our code will be made publicly available to encourage further research.

Large Language Models (LLMs) achieve impressive reasoning capabilities at the cost of substantial inference overhead, posing substantial deployment challenges. Although distilled Small Language Models (SLMs) significantly enhance efficiency, their performance suffers as they fail to follow LLMs' reasoning paths. Luckily, we reveal that only a small fraction of tokens genuinely diverge reasoning paths between LLMs and SLMs. Most generated tokens are either identical or exhibit neutral differences, such as minor variations in abbreviations or expressions. Leveraging this insight, we introduce **Roads to Rome (R2R)**, a neural token routing method that selectively utilizes LLMs only for these critical, path-divergent tokens, while leaving the majority of token generation to the SLM. We also develop an automatic data generation pipeline that identifies divergent tokens and generates token-level routing labels to train the lightweight router. We apply R2R to combine R1-1.5B and R1-32B models from the DeepSeek family, and evaluate on challenging math, coding, and QA benchmarks. With an average activated parameter size of 5.6B, R2R surpasses the average accuracy of R1-7B by 1.6x, outperforming even the R1-14B model. Compared to R1-32B, it delivers a 2.8x wall-clock speedup with comparable performance, advancing the Pareto frontier of test-time scaling efficiency. Our code is available at https://github.com/thu-nics/R2R.

The rapid development of Multimodal Large Language Models (MLLMs) has expanded their capabilities from image comprehension to video understanding. However, most of these MLLMs focus primarily on offline video comprehension, necessitating extensive processing of all video frames before any queries can be made. This presents a significant gap compared to the human ability to watch, listen, think, and respond to streaming inputs in real time, highlighting the limitations of current MLLMs. In this paper, we introduce StreamingBench, the first comprehensive benchmark designed to evaluate the streaming video understanding capabilities of MLLMs. StreamingBench assesses three core aspects of streaming video understanding: (1) real-time visual understanding, (2) omni-source understanding, and (3) contextual understanding. The benchmark consists of 18 tasks, featuring 900 videos and 4,500 human-curated QA pairs. Each video features five questions presented at different time points to simulate a continuous streaming scenario. We conduct experiments on StreamingBench with 13 open-source and proprietary MLLMs and find that even the most advanced proprietary MLLMs like Gemini 1.5 Pro and GPT-4o perform significantly below human-level streaming video understanding capabilities. We hope our work can facilitate further advancements for MLLMs, empowering them to approach human-level video comprehension and interaction in more realistic scenarios.

The recent exponential growth of Large Language Models (LLMs) has relied on GPU-based systems. However, CPUs are emerging as a flexible and lower-cost alternative, especially when targeting inference and reasoning workloads. RISC-V is rapidly gaining traction in this area, given its open and vendor-neutral ISA. However, the RISC-V hardware for LLM workloads and the corresponding software ecosystem are not fully mature and streamlined, given the requirement of domain-specific tuning. This paper aims at filling this gap, focusing on optimizing LLM inference on the Sophon SG2042, the first commercially available many-core RISC-V CPU with vector processing capabilities. On two recent state-of-the-art LLMs optimized for reasoning, DeepSeek R1 Distill Llama 8B and DeepSeek R1 Distill QWEN 14B, we achieve 4.32/2.29 token/s for token generation and 6.54/3.68 token/s for prompt processing, with a speed up of up 2.9x/3.0x compared to our baseline.

Routing large language models (LLMs) is a new paradigm that uses a router to recommend the best LLM from a pool of candidates for a given input. In this paper, our comprehensive analysis with more than 8,500 LLMs reveals a novel model-level scaling up phenomenon in Routing LLMs, i.e., a capable router can significantly enhance the performance of this paradigm as the number of candidates increases. This improvement can even surpass the performance of the best single model in the pool and many existing strong LLMs, confirming it a highly promising paradigm. However, the lack of comprehensive and open-source benchmarks for Routing LLMs has hindered the development of routers. In this paper, we introduce RouterEval, a benchmark tailored for router research, which includes over 200,000,000 performance records for 12 popular LLM evaluations across various areas such as commonsense reasoning, semantic understanding, etc., based on over 8,500 various LLMs. Using RouterEval, extensive evaluations of existing Routing LLM methods reveal that most still have significant room for improvement. See https://github.com/MilkThink-Lab/RouterEval for all data, code and tutorial.

Protecting intellectual property (IP) of text such as articles and code is increasingly important, especially as sophisticated attacks become possible, such as paraphrasing by large language models (LLMs) or even unauthorized training of LLMs on copyrighted text to infringe such IP. However, existing text watermarking methods are not robust enough against such attacks nor scalable to millions of users for practical implementation. In this paper, we propose Waterfall, the first training-free framework for robust and scalable text watermarking applicable across multiple text types (e.g., articles, code) and languages supportable by LLMs, for general text and LLM data provenance. Waterfall comprises several key innovations, such as being the first to use LLM as paraphrasers for watermarking along with a novel combination of techniques that are surprisingly effective in achieving robust verifiability and scalability. We empirically demonstrate that Waterfall achieves significantly better scalability, robust verifiability, and computational efficiency compared to SOTA article-text watermarking methods, and showed how it could be directly applied to the watermarking of code. We also demonstrated that Waterfall can be used for LLM data provenance, where the watermarks of LLM training data can be detected in LLM output, allowing for detection of unauthorized use of data for LLM training and potentially enabling model-centric watermarking of open-sourced LLMs which has been a limitation of existing LLM watermarking works. Our code is available at https://github.com/aoi3142/Waterfall.

It is desirable but challenging to generate content-rich long videos in the scale of minutes. Autoregressive large language models (LLMs) have achieved great success in generating coherent and long sequences of tokens in the domain of natural language processing, while the exploration of autoregressive LLMs for video generation is limited to generating short videos of several seconds. In this work, we conduct a deep analysis of the challenges that prevent autoregressive LLM-based video generators from generating long videos. Based on the observations and analysis, we propose Loong, a new autoregressive LLM-based video generator that can generate minute-long videos. Specifically, we model the text tokens and video tokens as a unified sequence for autoregressive LLMs and train the model from scratch. We propose progressive short-to-long training with a loss re-weighting scheme to mitigate the loss imbalance problem for long video training. We further investigate inference strategies, including video token re-encoding and sampling strategies, to diminish error accumulation during inference. Our proposed Loong can be trained on 10-second videos and be extended to generate minute-level long videos conditioned on text prompts, as demonstrated by the results. More samples are available at: https://epiphqny.github.io/Loong-video.

Large Language Models (LLMs) for complex reasoning is often hindered by high computational costs and latency, while resource-efficient Small Language Models (SLMs) typically lack the necessary reasoning capacity. Existing collaborative approaches, such as cascading or routing, operate at a coarse granularity by offloading entire queries to LLMs, resulting in significant computational waste when the SLM is capable of handling the majority of reasoning steps. To address this, we propose RelayLLM, a novel framework for efficient reasoning via token-level collaborative decoding. Unlike routers, RelayLLM empowers the SLM to act as an active controller that dynamically invokes the LLM only for critical tokens via a special command, effectively "relaying" the generation process. We introduce a two-stage training framework, including warm-up and Group Relative Policy Optimization (GRPO) to teach the model to balance independence with strategic help-seeking. Empirical results across six benchmarks demonstrate that RelayLLM achieves an average accuracy of 49.52%, effectively bridging the performance gap between the two models. Notably, this is achieved by invoking the LLM for only 1.07% of the total generated tokens, offering a 98.2% cost reduction compared to performance-matched random routers.

Decoder-only Large Language Models (LLMs) are currently the model of choice for many Natural Language Processing (NLP) applications. Through instruction fine-tuning and prompting approaches, such LLMs have been efficiently used to solve both general and domain-specific tasks. However, they are costly to train and, to a certain extent, costly to use as well, and one can wonder whether LLMs can be replaced by domain-specific Tiny Language Models (TLMs), which typically contain less than 100M parameters. We address this question in this study by comparing the performance of an 88M TLM pretrained from scratch for 185 A100 hours on a specific domain with a domain-specific tokenizer (here, the Indian legal domain) with LLMs of various sizes between 1B and 8B for solving domain-specific tasks. We show in particular that our legal TLM, Ayn, can indeed outperform LLMs up to 80 times larger on the legal case judgment prediction task, rival LLMs up to 30 times larger on the summarization task, and still be competitive with these larger LLMs on general tasks.

The increasing context window size in Large Language Models (LLMs), such as the GPT and LLaMA series, has improved their ability to tackle complex, long-text tasks, but at the cost of inference efficiency, particularly regarding memory and computational complexity. Existing methods, including selective token retention and window-based attention, improve efficiency but risk discarding important tokens needed for future text generation. In this paper, we propose an approach that enhances LLM efficiency without token loss by reducing the memory and computational load of less important tokens, rather than discarding them.We address two challenges: 1) investigating the distribution of important tokens in the context, discovering recent tokens are more important than distant tokens in context, and 2) optimizing resources for distant tokens by sharing attention scores across layers. The experiments show that our method saves 35% KV cache without compromising the performance.

Advancements in Large Language Models (LLMs) inspire various strategies for integrating video modalities. A key approach is Video-LLMs, which incorporate an optimizable interface linking sophisticated video encoders to LLMs. However, due to computation and data limitations, these Video-LLMs are typically pre-trained to process only short videos, limiting their broader application for understanding longer video content. Additionally, fine-tuning Video-LLMs to handle longer videos is cost-prohibitive. Consequently, it becomes essential to explore the interpolation of Video-LLMs under a completely training-free setting. In this paper, we first identify the primary challenges in interpolating Video-LLMs: (1) the video encoder and modality alignment projector are fixed, preventing the integration of additional frames into Video-LLMs, and (2) the LLM backbone is limited in its content length capabilities, which complicates the processing of an increased number of video tokens. To address these challenges, we propose a specific INTerPolation method for Video-LLMs (INTP-Video-LLMs). We introduce an alternative video token rearrangement technique that circumvents limitations imposed by the fixed video encoder and alignment projector. Furthermore, we introduce a training-free LLM context window extension method to enable Video-LLMs to understand a correspondingly increased number of visual tokens.

Increasing demand for Large Language Models (LLMs) services imposes substantial deployment and computation costs on providers. LLM routing offers a cost-efficient solution by directing queries to the optimal LLM based on model and query features. However, existing works primarily focus on offline scenarios and struggle to adapt to online settings with high query volume and constrained token budgets. In this work, we introduce the first training-free algorithm for online routing scenarios. Our algorithm leverages approximate nearest neighbor search to efficiently estimate query features and performs a one-time optimization over a small set of initial queries to learn a routing strategy that guides future routing. We provide theoretical guarantees demonstrating that our algorithm achieves a competitive ratio of 1 - o(1) under natural assumptions, which is further validated by extensive experiments across 3 benchmark datasets and 8 baselines, showing an average improvement of 3.55times in overall performance, 1.85times in cost efficiency, and nearly 4.25times in throughput. Our code is available at https://github.com/fzwark/PORT.

Large language models (LLMs) have emerged as a cornerstone in real-world applications with lengthy streaming inputs, such as LLM-driven agents. However, existing LLMs, pre-trained on sequences with restricted maximum length, cannot generalize to longer sequences due to the out-of-domain and distraction issues. To alleviate these issues, existing efforts employ sliding attention windows and discard distant tokens to achieve the processing of extremely long sequences. Unfortunately, these approaches inevitably fail to capture long-distance dependencies within sequences to deeply understand semantics. This paper introduces a training-free memory-based method, InfLLM, to unveil the intrinsic ability of LLMs to process streaming long sequences. Specifically, InfLLM stores distant contexts into additional memory units and employs an efficient mechanism to lookup token-relevant units for attention computation. Thereby, InfLLM allows LLMs to efficiently process long sequences while maintaining the ability to capture long-distance dependencies. Without any training, InfLLM enables LLMs pre-trained on sequences of a few thousand tokens to achieve superior performance than competitive baselines continually training these LLMs on long sequences. Even when the sequence length is scaled to 1,024K, InfLLM still effectively captures long-distance dependencies.

Large language models (LLMs) are useful in many NLP tasks and become more capable with size, with the best open-source models having over 50 billion parameters. However, using these 50B+ models requires high-end hardware, making them inaccessible to most researchers. In this work, we investigate methods for cost-efficient inference and fine-tuning of LLMs, comparing local and distributed strategies. We observe that a large enough model (50B+) can run efficiently even on geodistributed devices in a consumer-grade network. This could allow running LLM efficiently by pooling together idle compute resources of multiple research groups and volunteers. We address two open problems: (1) how to perform inference and fine-tuning reliably if any device can disconnect abruptly and (2) how to partition LLMs between devices with uneven hardware, joining and leaving at will. In order to do that, we develop special fault-tolerant inference algorithms and load-balancing protocols that automatically assign devices to maximize the total system throughput. We showcase these algorithms in Petals - a decentralized system that runs Llama 2 (70B) and BLOOM (176B) over the Internet up to 10x faster than offloading for interactive generation. We evaluate the performance of our system in simulated conditions and a real-world setup spanning two continents.

Large Language Models (LLMs) have been widely used in various tasks, motivating us to develop an LLM-based assistant for videos. Instead of training from scratch, we propose a module to transform arbitrary well-trained image-based LLMs into video-LLMs (after being trained on video data). To better adapt image-LLMs for processing videos, we introduce two design principles: linear transformation to preserve the original visual-language alignment and representative information condensation from redundant video content. Guided by these principles, we propose a plug-and-play Linear Video Tokenizer(LinVT), which enables existing image-LLMs to understand videos. We benchmark LinVT with six recent visual LLMs: Aquila, Blip-3, InternVL2, Mipha, Molmo and Qwen2-VL, showcasing the high compatibility of LinVT. LinVT-based LLMs achieve state-of-the-art performance across various video benchmarks, illustrating the effectiveness of LinVT in multi-modal video understanding.

Practical large language model (LLM) services may involve a long system prompt, which specifies the instructions, examples, and knowledge documents of the task and is reused across numerous requests. However, the long system prompt causes throughput/latency bottlenecks as the cost of generating the next token grows w.r.t. the sequence length. This paper aims to improve the efficiency of LLM services that involve long system prompts. Our key observation is that handling these system prompts requires heavily redundant memory accesses in existing causal attention computation algorithms. Specifically, for batched requests, the cached hidden states (i.e., key-value pairs) of system prompts are transferred from off-chip DRAM to on-chip SRAM multiple times, each corresponding to an individual request. To eliminate such a redundancy, we propose RelayAttention, an attention algorithm that allows reading these hidden states from DRAM exactly once for a batch of input tokens. RelayAttention is a free lunch: it maintains the generation quality while requiring no model retraining, as it is based on a mathematical reformulation of causal attention.

When the complete source sentence is provided, Large Language Models (LLMs) perform excellently in offline machine translation even with a simple prompt "Translate the following sentence from [src lang] into [tgt lang]:". However, in many real scenarios, the source tokens arrive in a streaming manner and simultaneous machine translation (SiMT) is required, then the efficiency and performance of decoder-only LLMs are significantly limited by their auto-regressive nature. To enable LLMs to achieve high-quality SiMT as efficiently as offline translation, we propose a novel paradigm that includes constructing supervised fine-tuning (SFT) data for SiMT, along with new training and inference strategies. To replicate the token input/output stream in SiMT, the source and target tokens are rearranged into an interleaved sequence, separated by special tokens according to varying latency requirements. This enables powerful LLMs to learn read and write operations adaptively, based on varying latency prompts, while still maintaining efficient auto-regressive decoding. Experimental results show that, even with limited SFT data, our approach achieves state-of-the-art performance across various SiMT benchmarks, and preserves the original abilities of offline translation. Moreover, our approach generalizes well to document-level SiMT setting without requiring specific fine-tuning, even beyond the offline translation model.

Large language models (LLMs) power a new generation of interactive AI applications exemplified by ChatGPT. The interactive nature of these applications demand low job completion time (JCT) for model inference. Existing LLM serving systems use run-to-completion processing for inference jobs, which suffers from head-of-line blocking and long JCT. We present FastServe, a distributed inference serving system for LLMs. FastServe exploits the autoregressive pattern of LLM inference to enable preemption at the granularity of each output token. FastServe uses preemptive scheduling to minimize JCT with a novel skip-join Multi-Level Feedback Queue scheduler. Based on the new semi information-agnostic setting of LLM inference, the scheduler leverages the input length information to assign an appropriate initial queue for each arrival job to join. The higher priority queues than the joined queue are skipped to reduce demotions. We design an efficient GPU memory management mechanism that proactively offloads and uploads intermediate states between GPU memory and host memory for LLM inference. We build a system prototype of FastServe based on NVIDIA FasterTransformer. Experimental results show that compared to the state-of-the-art solution Orca, FastServe improves the average and tail JCT by up to 5.1times and 6.4times, respectively.

Transformer-based large language model (LLM) inference serving is now the backbone of many cloud services. LLM inference consists of a prefill phase and a decode phase. However, existing LLM deployment practices often overlook the distinct characteristics of these phases, leading to significant interference. To mitigate interference, our insight is to carefully schedule and group inference requests based on their characteristics. We realize this idea in TetriInfer through three pillars. First, it partitions prompts into fixed-size chunks so that the accelerator always runs close to its computationsaturated limit. Second, it disaggregates prefill and decode instances so each can run independently. Finally, it uses a smart two-level scheduling algorithm augmented with predicted resource usage to avoid decode scheduling hotspots. Results show that TetriInfer improves time-to-first-token (TTFT), job completion time (JCT), and inference efficiency in turns of performance per dollar by a large margin, e.g., it uses 38% less resources all the while lowering average TTFT and average JCT by 97% and 47%, respectively.

Large language models (LLMs) are increasingly deployed in specialized production data processing pipelines across diverse domains -- such as finance, marketing, and e-commerce. However, when running them in production across many inputs, they often fail to follow instructions or meet developer expectations. To improve reliability in these applications, creating assertions or guardrails for LLM outputs to run alongside the pipelines is essential. Yet, determining the right set of assertions that capture developer requirements for a task is challenging. In this paper, we introduce PROMPTEVALS, a dataset of 2087 LLM pipeline prompts with 12623 corresponding assertion criteria, sourced from developers using our open-source LLM pipeline tools. This dataset is 5x larger than previous collections. Using a hold-out test split of PROMPTEVALS as a benchmark, we evaluated closed- and open-source models in generating relevant assertions. Notably, our fine-tuned Mistral and Llama 3 models outperform GPT-4o by 20.93% on average, offering both reduced latency and improved performance. We believe our dataset can spur further research in LLM reliability, alignment, and prompt engineering.

The Large Language Model (LLM) watermark is a newly emerging technique that shows promise in addressing concerns surrounding LLM copyright, monitoring AI-generated text, and preventing its misuse. The LLM watermark scheme commonly includes generating secret keys to partition the vocabulary into green and red lists, applying a perturbation to the logits of tokens in the green list to increase their sampling likelihood, thus facilitating watermark detection to identify AI-generated text if the proportion of green tokens exceeds a threshold. However, recent research indicates that watermarking methods using numerous keys are susceptible to removal attacks, such as token editing, synonym substitution, and paraphrasing, with robustness declining as the number of keys increases. Therefore, the state-of-the-art watermark schemes that employ fewer or single keys have been demonstrated to be more robust against text editing and paraphrasing. In this paper, we propose a novel green list stealing attack against the state-of-the-art LLM watermark scheme and systematically examine its vulnerability to this attack. We formalize the attack as a mixed integer programming problem with constraints. We evaluate our attack under a comprehensive threat model, including an extreme scenario where the attacker has no prior knowledge, lacks access to the watermark detector API, and possesses no information about the LLM's parameter settings or watermark injection/detection scheme. Extensive experiments on LLMs, such as OPT and LLaMA, demonstrate that our attack can successfully steal the green list and remove the watermark across all settings.

Advanced Large Language Models (LLMs) have achieved impressive performance across a wide range of complex and long-context natural language tasks. However, performing long-context LLM inference locally on a commodity GPU (a PC) with privacy concerns remains challenging due to the increasing memory demands of the key-value (KV) cache. Existing systems typically identify important tokens and selectively offload their KV data to GPU and CPU memory. The KV data needs to be offloaded to disk due to the limited memory on a commodity GPU, but the process is bottlenecked by token importance evaluation overhead and the disk's low bandwidth. In this paper, we present LeoAM, the first efficient importance-aware long-context LLM inference system for a single commodity GPU with adaptive hierarchical GPU-CPU-Disk KV management. Our system employs an adaptive KV management strategy that partitions KV data into variable-sized chunks based on the skewed distribution of attention weights across different layers to reduce computational and additional transmission overheads. Moreover, we propose a lightweight KV abstract method, which minimizes transmission latency by storing and extracting the KV abstract of each chunk on disk instead of the full KV data. LeoAM also leverages the dynamic compression and pipeline techniques to further accelerate inference. Experimental results demonstrate that LongInfer achieves an average inference latency speedup of 3.46x, while maintaining comparable LLM response quality. In scenarios with larger batch sizes, it achieves up to a 5.47x speedup.

Realistic evaluation of LLM serving systems requires online workloads, dynamic arrivals, queueing, and the serving engine's local scheduling for execution batching, but running such experiments on GPUs is expensive. Existing simulators reduce this cost, but often operate offline or in time-warped mode, re-implement serving-engine schedulers, or require accurate operator/kernel-level latency models. We present LLM-Emu, a serving-native emulator for vLLM that preserves the production HTTP, scheduling, KV-cache, and output-processing paths while replacing only GPU forward execution with profile-sampled latency and synthetic output tokens. Tested on two different GPUs, four model variants, two model families, two attention backends, and both Poisson and bursty ShareGPT workloads, LLM-Emu closely tracks real vLLM serving behavior: TPOT and ITL stay within 4.8% absolute error, E2E latency within 5.3%, and output throughput within 1.9%; TTFT is less stable, with maximum error 10.4%, reflecting its sensitivity to admission and queue state. These results suggest that lightweight, serving-native emulation can support practical online experimentation for LLM-serving systems. LLM-Emu is open sourced at https://github.com/AKafakA/llm-emu.

The fundamental challenge in scaling Video Large Language Models (Video LLMs) to long-form video lies in managing the explosion of visual-token context length. Existing strategies predominantly focus on "post-hoc" token reduction -- reducing visual tokens after feature extraction to alleviate the LLM's computational overhead. While these methods effectively reduce the number of visual tokens, we observe that the primary latency bottleneck then shifts from the LLM to the expensive per-frame processing of the vision encoder. To address this, we introduce LiteFrame, a strong, yet highly efficient video encoder backbone for Video LLMs. To train LiteFrame, we propose Compressed Token Distillation (CTD), a novel training framework that teaches a compact student vision encoder to directly predict information-dense, spatio-temporally compressed representations produced by a large teacher vision model, effectively bypassing redundant computation. When coupled with further Language Model Adaptation (LMA), this approach results in a new latency-accuracy Pareto frontier -- compared with InternVL3-8B, LiteFrame provides a 35% reduction in end-to-end latency while processing 8times more frames and improves average video understanding accuracy across multiple benchmarks. Our results demonstrate a new potential path to unlocking longer-form video understanding under fixed compute budgets.

Large Language Models (LLMs), such as ChatGPT, LLaMA, GLM, and PaLM, have exhibited remarkable performances across various tasks in recent years. However, LLMs face two main challenges in real-world applications. One challenge is that training LLMs consumes vast computing resources, preventing LLMs from being adopted by small and medium-sized enterprises with limited computing resources. Another is that training LLM requires a large amount of high-quality data, which are often scattered among enterprises. To address these challenges, we propose FATE-LLM, an industrial-grade federated learning framework for large language models. FATE-LLM (1) facilitates federated learning for large language models (coined FedLLM); (2) promotes efficient training of FedLLM using parameter-efficient fine-tuning methods; (3) protects the intellectual property of LLMs; (4) preserves data privacy during training and inference through privacy-preserving mechanisms. We release the code of FATE-LLM at https://github.com/FederatedAI/FATE-LLM to facilitate the research of FedLLM and enable a broad range of industrial applications.

In the development of Large Language Models (LLMs), considerable attention has been given to the quality of training datasets. However, the role of tokenizers in the LLM training pipeline, particularly for multilingual models, has received less focus. The quality of tokenization can significantly impact a model's ability to handle diverse languages effectively. We introduce Qtok, a tool designed to assess tokenizer quality with a specific emphasis on their performance in multilingual contexts. Our research proposes a set of metrics for evaluating tokenizer quality, including measures of language coverage, token completeness, and distribution across languages and linguistic categories. Qtok applies these metrics to evaluate 13 distinct tokenizers from 58 publicly available models, analyzing their output across different linguistic contexts. Our analysis revealed significant variations in token distribution across languages and categories, highlighting potential biases and areas for improvement in current tokenization strategies. This research contributes to the field of tokenizer evaluation within multilingual LLM development by providing a systematic approach to assessing tokenizer quality. Our findings highlight the critical role of tokenization in multilingual LLM capability. The Qtok tool and our analysis methodology offer practical means for researchers to evaluate and improve tokenization strategies for multilingual applications. We offer a method to compare tokenizer quality across these metrics, which may be useful when selecting or adjusting tokenizers for specific multilingual LLM applications.

The rapid proliferation of Large Language Models (LLMs) has been a driving force in the growth of cloud-based LLM services, which are now integral to advancing AI applications. However, the dynamic auto-regressive nature of LLM service, along with the need to support exceptionally long context lengths, demands the flexible allocation and release of substantial resources. This presents considerable challenges in designing cloud-based LLM service systems, where inefficient management can lead to performance degradation or resource wastage. In response to these challenges, this work introduces DistAttention, a novel distributed attention algorithm that segments the KV Cache into smaller, manageable units, enabling distributed processing and storage of the attention module. Based on that, we propose DistKV-LLM, a distributed LLM serving system that dynamically manages KV Cache and effectively orchestrates all accessible GPU and CPU memories spanning across the data center. This ensures a high-performance LLM service on the cloud, adaptable to a broad range of context lengths. Validated in a cloud environment with 32 NVIDIA A100 GPUs in configurations from 2 to 32 instances, our system exhibited 1.03-2.4x end-to-end throughput improvements and supported context lengths 2-19x longer than current state-of-the-art LLM service systems, as evidenced by extensive testing across 18 datasets with context lengths up to 1,900K.

LLM inference for popular enterprise use cases, such as summarization, RAG, and code-generation, typically observes orders of magnitude longer prompt lengths than generation lengths. This characteristic leads to high cost of prefill and increased response latency. In this paper, we present SwiftKV, a novel model transformation and distillation procedure specifically designed to reduce the time and cost of processing prompt tokens while preserving high quality of generated tokens. SwiftKV combines three key mechanisms: i) SingleInputKV, which prefills later layers' KV cache using a much earlier layer's output, allowing prompt tokens to skip much of the model computation, ii) AcrossKV, which merges the KV caches of neighboring layers to reduce the memory footprint and support larger batch size for higher throughput, and iii) a knowledge-preserving distillation procedure that can adapt existing LLMs for SwiftKV with minimal accuracy impact and low compute and data requirement. For Llama-3.1-8B and 70B, SwiftKV reduces the compute requirement of prefill by 50% and the memory requirement of the KV cache by 62.5% while incurring minimum quality degradation across a wide range of tasks. In the end-to-end inference serving using an optimized vLLM implementation, SwiftKV realizes up to 2x higher aggregate throughput and 60% lower time per output token. It can achieve a staggering 560 TFlops/GPU of normalized inference throughput, which translates to 16K tokens/s for Llama-3.1-70B in 16-bit precision on 4x H100 GPUs.

The Large Language Model (LLM) is widely employed for tasks such as intelligent assistants, text summarization, translation, and multi-modality on mobile phones. However, the current methods for on-device LLM deployment maintain slow inference speed, which causes poor user experience. To facilitate high-efficiency LLM deployment on device GPUs, we propose four optimization techniques: (a) a symbolic expression-based approach to support dynamic shape model inference; (b) operator optimizations and execution priority setting to enhance inference speed and reduce phone lagging; (c) an FP4 quantization method termed M0E4 to reduce dequantization overhead; (d) a sub-tensor-based technique to eliminate the need for copying KV cache after LLM inference. Furthermore, we implement these methods in our mobile inference engine, Transformer-Lite, which is compatible with both Qualcomm and MTK processors. We evaluated Transformer-Lite's performance using LLMs with varied architectures and parameters ranging from 2B to 14B. Specifically, we achieved prefill and decoding speeds of 121 token/s and 14 token/s for ChatGLM2 6B, and 330 token/s and 30 token/s for smaller Gemma 2B, respectively. Compared with CPU-based FastLLM and GPU-based MLC-LLM, our engine attains over 10x speedup for the prefill speed and 2~3x speedup for the decoding speed.

LLM-based Multi-Agent (LLM-MA) systems are increasingly applied to automate complex software engineering tasks such as requirements engineering, code generation, and testing. However, their operational efficiency and resource consumption remain poorly understood, hindering practical adoption due to unpredictable costs and environmental impact. To address this, we conduct an analysis of token consumption patterns in an LLM-MA system within the Software Development Life Cycle (SDLC), aiming to understand where tokens are consumed across distinct software engineering activities. We analyze execution traces from 30 software development tasks performed by the ChatDev framework using a GPT-5 reasoning model, mapping its internal phases to distinct development stages (Design, Coding, Code Completion, Code Review, Testing, and Documentation) to create a standardized evaluation framework. We then quantify and compare token distribution (input, output, reasoning) across these stages. Our preliminary findings show that the iterative Code Review stage accounts for the majority of token consumption for an average of 59.4% of tokens. Furthermore, we observe that input tokens consistently constitute the largest share of consumption for an average of 53.9%, providing empirical evidence for potentially significant inefficiencies in agentic collaboration. Our results suggest that the primary cost of agentic software engineering lies not in initial code generation but in automated refinement and verification. Our novel methodology can help practitioners predict expenses and optimize workflows, and it directs future research toward developing more token-efficient agent collaboration protocols.

Recent advances in Large Language Models (LLMs) have enabled the development of Video-LLMs, advancing multimodal learning by bridging video data with language tasks. However, current video understanding models struggle with processing long video sequences, supporting multi-turn dialogues, and adapting to real-world dynamic scenarios. To address these issues, we propose StreamChat, a training-free framework for streaming video reasoning and conversational interaction. StreamChat leverages a novel hierarchical memory system to efficiently process and compress video features over extended sequences, enabling real-time, multi-turn dialogue. Our framework incorporates a parallel system scheduling strategy that enhances processing speed and reduces latency, ensuring robust performance in real-world applications. Furthermore, we introduce StreamBench, a versatile benchmark that evaluates streaming video understanding across diverse media types and interactive scenarios, including multi-turn interactions and complex reasoning tasks. Extensive evaluations on StreamBench and other public benchmarks demonstrate that StreamChat significantly outperforms existing state-of-the-art models in terms of accuracy and response times, confirming its effectiveness for streaming video understanding. Code is available at StreamChat: https://github.com/hmxiong/StreamChat.

Inference serving for large language models (LLMs) is the key to unleashing their potential in people's daily lives. However, efficient LLM serving remains challenging today because the requests are inherently heterogeneous and unpredictable in terms of resource and latency requirements, as a result of the diverse applications and the dynamic execution nature of LLMs. Existing systems are fundamentally limited in handling these characteristics and cause problems such as severe queuing delays, poor tail latencies, and SLO violations. We introduce Llumnix, an LLM serving system that reacts to such heterogeneous and unpredictable requests by runtime rescheduling across multiple model instances. Similar to context switching across CPU cores in modern operating systems, Llumnix reschedules requests to improve load balancing and isolation, mitigate resource fragmentation, and differentiate request priorities and SLOs. Llumnix implements the rescheduling with an efficient and scalable live migration mechanism for requests and their in-memory states, and exploits it in a dynamic scheduling policy that unifies the multiple rescheduling scenarios elegantly. Our evaluations show that Llumnix improves tail latencies by an order of magnitude, accelerates high-priority requests by up to 1.5x, and delivers up to 36% cost savings while achieving similar tail latencies, compared against state-of-the-art LLM serving systems. Llumnix is publicly available at https://github.com/AlibabaPAI/llumnix.

The rapid growth of online video platforms, particularly live streaming services, has created an urgent need for real-time video understanding systems. These systems must process continuous video streams and respond to user queries instantaneously, presenting unique challenges for current Video Large Language Models (VideoLLMs). While existing VideoLLMs excel at processing complete videos, they face significant limitations in streaming scenarios due to their inability to handle dense, redundant frames efficiently. We introduce TimeChat-Online, a novel online VideoLLM that revolutionizes real-time video interaction. At its core lies our innovative Differential Token Drop (DTD) module, which addresses the fundamental challenge of visual redundancy in streaming videos. Drawing inspiration from human visual perception's Change Blindness phenomenon, DTD preserves meaningful temporal changes while filtering out static, redundant content between frames. Remarkably, our experiments demonstrate that DTD achieves an 82.8% reduction in video tokens while maintaining 98% performance on StreamingBench, revealing that over 80% of visual content in streaming videos is naturally redundant without requiring language guidance. To enable seamless real-time interaction, we present TimeChat-Online-139K, a comprehensive streaming video dataset featuring diverse interaction patterns including backward-tracing, current-perception, and future-responding scenarios. TimeChat-Online's unique Proactive Response capability, naturally achieved through continuous monitoring of video scene transitions via DTD, sets it apart from conventional approaches. Our extensive evaluation demonstrates TimeChat-Online's superior performance on streaming benchmarks (StreamingBench and OvOBench) and maintaining competitive results on long-form video tasks such as Video-MME and MLVU.

Frontier LLMs are increasingly utilised across academia, society and industry. A commonly used unit for comparing models, their inputs and outputs, and estimating inference pricing is the token. In general, tokens are used as a stable currency, assumed to be broadly consistent across tokenizers and contexts, enabling direct comparisons. However, tokenization varies significantly across models and domains of text, making naive interpretation of token counts problematic. We quantify this variation by providing a comprehensive empirical analysis of tokenization, exploring the compression of sequences to tokens across different distributions of textual data. Our analysis challenges commonly held heuristics about token lengths, finding them to be overly simplistic. We hope the insights of our study add clarity and intuition toward tokenization in contemporary LLMs.

In the wave of generative recommendation, we present OneMall, an end-to-end generative recommendation framework tailored for e-commerce services at Kuaishou. Our OneMall systematically unifies the e-commerce's multiple item distribution scenarios, such as Product-card, short-video and live-streaming. Specifically, it comprises three key components, aligning the entire model training pipeline to the LLM's pre-training/post-training: (1) E-commerce Semantic Tokenizer: we provide a tokenizer solution that captures both real-world semantics and business-specific item relations across different scenarios; (2) Transformer-based Architecture: we largely utilize Transformer as our model backbone, e.g., employing Query-Former for long sequence compression, Cross-Attention for multi-behavior sequence fusion, and Sparse MoE for scalable auto-regressive generation; (3) Reinforcement Learning Pipeline: we further connect retrieval and ranking models via RL, enabling the ranking model to serve as a reward signal for end-to-end policy retrieval model optimization. Extensive experiments demonstrate that OneMall achieves consistent improvements across all e-commerce scenarios: +13.01\% GMV in product-card, +15.32\% Orders in Short-Video, and +2.78\% Orders in Live-Streaming. OneMall has been deployed, serving over 400 million daily active users at Kuaishou.

Transformers, driven by attention mechanisms, form the foundation of large language models (LLMs). As these models scale up, efficient GPU attention kernels become essential for high-throughput and low-latency inference. Diverse LLM applications demand flexible and high-performance attention solutions. We present FlashInfer: a customizable and efficient attention engine for LLM serving. FlashInfer tackles KV-cache storage heterogeneity using block-sparse format and composable formats to optimize memory access and reduce redundancy. It also offers a customizable attention template, enabling adaptation to various settings through Just-In-Time (JIT) compilation. Additionally, FlashInfer's load-balanced scheduling algorithm adjusts to dynamism of user requests while maintaining compatibility with CUDAGraph which requires static configuration. FlashInfer have been integrated into leading LLM serving frameworks like SGLang, vLLM and MLC-Engine. Comprehensive kernel-level and end-to-end evaluations demonstrate FlashInfer's ability to significantly boost kernel performance across diverse inference scenarios: compared to state-of-the-art LLM serving solutions, FlashInfer achieve 29-69% inter-token-latency reduction compared to compiler backends for LLM serving benchmark, 28-30% latency reduction for long-context inference, and 13-17% speedup for LLM serving with parallel generation.

A well-known dilemma in large vision-language models (e.g., GPT-4, LLaVA) is that while increasing the number of vision tokens generally enhances visual understanding, it also significantly raises memory and computational costs, especially in long-term, dense video frame streaming scenarios. Although learnable approaches like Q-Former and Perceiver Resampler have been developed to reduce the vision token burden, they overlook the context causally modeled by LLMs (i.e., key-value cache), potentially leading to missed visual cues when addressing user queries. In this paper, we introduce a novel approach to reduce vision compute by leveraging redundant vision tokens "skipping layers" rather than decreasing the number of vision tokens. Our method, VideoLLM-MoD, is inspired by mixture-of-depths LLMs and addresses the challenge of numerous vision tokens in long-term or streaming video. Specifically, for each transformer layer, we learn to skip the computation for a high proportion (e.g., 80\%) of vision tokens, passing them directly to the next layer. This approach significantly enhances model efficiency, achieving approximately \textasciitilde42\% time and \textasciitilde30\% memory savings for the entire training. Moreover, our method reduces the computation in the context and avoid decreasing the vision tokens, thus preserving or even improving performance compared to the vanilla model. We conduct extensive experiments to demonstrate the effectiveness of VideoLLM-MoD, showing its state-of-the-art results on multiple benchmarks, including narration, forecasting, and summarization tasks in COIN, Ego4D, and Ego-Exo4D datasets.

Recent video large language models (Video LLMs) often depend on costly human annotations or proprietary model APIs (e.g., GPT-4o) to produce training data, which limits their training at scale. In this paper, we explore large-scale training for Video LLM with cheap automatic speech recognition (ASR) transcripts. Specifically, we propose a novel streaming training approach that densely interleaves the ASR words and video frames according to their timestamps. Compared to previous studies in vision-language representation with ASR, our method naturally fits the streaming characteristics of ASR, thus enabling the model to learn temporally-aligned, fine-grained vision-language modeling. To support the training algorithm, we introduce a data production pipeline to process YouTube videos and their closed captions (CC, same as ASR), resulting in Live-CC-5M dataset for pre-training and Live-WhisperX-526K dataset for high-quality supervised fine-tuning (SFT). Remarkably, even without SFT, the ASR-only pre-trained LiveCC-7B-Base model demonstrates competitive general video QA performance and exhibits a new capability in real-time video commentary. To evaluate this, we carefully design a new LiveSports-3K benchmark, using LLM-as-a-judge to measure the free-form commentary. Experiments show our final LiveCC-7B-Instruct model can surpass advanced 72B models (Qwen2.5-VL-72B-Instruct, LLaVA-Video-72B) in commentary quality even working in a real-time mode. Meanwhile, it achieves state-of-the-art results at the 7B/8B scale on popular video QA benchmarks such as VideoMME and OVOBench, demonstrating the broad generalizability of our approach. All resources of this paper have been released at https://showlab.github.io/livecc.

Multimodal large language models (MLLMs) have shown strong performance on offline video understanding, but most are limited to offline inference or have weak online reasoning, making multi-turn interaction over continuously arriving video streams difficult. Existing streaming methods typically use an interleaved perception-generation paradigm, which prevents concurrent perception and generation and leads to early memory decay as streams grow, hurting long-range dependency modeling. We propose Think While Watching, a memory-anchored streaming video reasoning framework that preserves continuous segment-level memory during multi-turn interaction. We build a three-stage, multi-round chain-of-thought dataset and adopt a stage-matched training strategy, while enforcing strict causality through a segment-level streaming causal mask and streaming positional encoding. During inference, we introduce an efficient pipeline that overlaps watching and thinking and adaptively selects the best attention backend. Under both single-round and multi-round streaming input protocols, our method achieves strong results. Built on Qwen3-VL, it improves single-round accuracy by 2.6% on StreamingBench and by 3.79% on OVO-Bench. In the multi-round setting, it maintains performance while reducing output tokens by 56%. Code is available at: https://github.com/wl666hhh/Think_While_Watching/

Large Language Models (LLMs) enable a new ecosystem with many downstream applications, called LLM applications, with different natural language processing tasks. The functionality and performance of an LLM application highly depend on its system prompt, which instructs the backend LLM on what task to perform. Therefore, an LLM application developer often keeps a system prompt confidential to protect its intellectual property. As a result, a natural attack, called prompt leaking, is to steal the system prompt from an LLM application, which compromises the developer's intellectual property. Existing prompt leaking attacks primarily rely on manually crafted queries, and thus achieve limited effectiveness. In this paper, we design a novel, closed-box prompt leaking attack framework, called PLeak, to optimize an adversarial query such that when the attacker sends it to a target LLM application, its response reveals its own system prompt. We formulate finding such an adversarial query as an optimization problem and solve it with a gradient-based method approximately. Our key idea is to break down the optimization goal by optimizing adversary queries for system prompts incrementally, i.e., starting from the first few tokens of each system prompt step by step until the entire length of the system prompt. We evaluate PLeak in both offline settings and for real-world LLM applications, e.g., those on Poe, a popular platform hosting such applications. Our results show that PLeak can effectively leak system prompts and significantly outperforms not only baselines that manually curate queries but also baselines with optimized queries that are modified and adapted from existing jailbreaking attacks. We responsibly reported the issues to Poe and are still waiting for their response. Our implementation is available at this repository: https://github.com/BHui97/PLeak.

We propose LLM-Interleaved (LLM-I), a flexible and dynamic framework that reframes interleaved image-text generation as a tool-use problem. LLM-I is designed to overcome the "one-tool" bottleneck of current unified models, which are limited to synthetic imagery and struggle with tasks requiring factual grounding or programmatic precision. Our framework empowers a central LLM or MLLM agent to intelligently orchestrate a diverse toolkit of specialized visual tools, including online image search, diffusion-based generation, code execution, and image editing. The agent is trained to select and apply these tools proficiently via a Reinforcement Learning (RL) framework that features a hybrid reward system combining rule-based logic with judgments from LLM and MLLM evaluators. Trained on a diverse new dataset using four different model backbones, LLM-I demonstrates state-of-the-art performance, outperforming existing methods by a large margin across four benchmarks. We also introduce a novel test-time scaling strategy that provides further performance gains. Project Page: https://github.com/ByteDance-BandAI/LLM-I.

The exploration and application of Large Language Models (LLMs) is thriving. To reduce deployment costs, continuous batching has become an essential feature in current service frameworks. The effectiveness of continuous batching relies on an accurate estimate of the memory requirements of requests. However, due to the diversity in request output lengths, existing frameworks tend to adopt aggressive or conservative schedulers, which often result in significant overestimation or underestimation of memory consumption. Consequently, they suffer from harmful request evictions or prolonged queuing times, failing to achieve satisfactory throughput under strict Service Level Agreement (SLA) guarantees (a.k.a. goodput), across various LLM application scenarios with differing input-output length distributions. To address this issue, we propose a novel Past-Future scheduler that precisely estimates the peak memory resources required by the running batch via considering the historical distribution of request output lengths and calculating memory occupancy at each future time point. It adapts to applications with all types of input-output length distributions, balancing the trade-off between request queuing and harmful evictions, thereby consistently achieving better goodput. Furthermore, to validate the effectiveness of the proposed scheduler, we developed a high-performance LLM serving framework, LightLLM, that implements the Past-Future scheduler. Compared to existing aggressive or conservative schedulers, LightLLM demonstrates superior goodput, achieving up to 2-3times higher goodput than other schedulers under heavy loads. LightLLM is open source to boost the research in such direction (https://github.com/ModelTC/lightllm).

Multimodal LLMs (MLLMs) are the natural extension of large language models to handle multimodal inputs, combining text and image data. They have recently garnered attention due to their capability to address complex tasks involving both modalities. However, their effectiveness is limited to the knowledge acquired during training, which restricts their practical utility. In this work, we introduce a novel method to enhance the adaptability of MLLMs by integrating external knowledge sources. Our proposed model, Reflective LLaVA (ReflectiVA), utilizes reflective tokens to dynamically determine the need for external knowledge and predict the relevance of information retrieved from an external database. Tokens are trained following a two-stage two-model training recipe. This ultimately enables the MLLM to manage external knowledge while preserving fluency and performance on tasks where external knowledge is not needed. Through our experiments, we demonstrate the efficacy of ReflectiVA for knowledge-based visual question answering, highlighting its superior performance compared to existing methods. Source code and trained models are publicly available at https://github.com/aimagelab/ReflectiVA.

Shannon, in his seminal paper introducing information theory, divided the communication into three levels: technical, semantic, and effectivenss. While the technical level is concerned with accurate reconstruction of transmitted symbols, the semantic and effectiveness levels deal with the inferred meaning and its effect on the receiver. Thanks to telecommunications, the first level problem has produced great advances like the internet. Large Language Models (LLMs) make some progress towards the second goal, but the third level still remains largely untouched. The third problem deals with predicting and optimizing communication for desired receiver behavior. LLMs, while showing wide generalization capabilities across a wide range of tasks, are unable to solve for this. One reason for the underperformance could be a lack of "behavior tokens" in LLMs' training corpora. Behavior tokens define receiver behavior over a communication, such as shares, likes, clicks, purchases, retweets, etc. While preprocessing data for LLM training, behavior tokens are often removed from the corpora as noise. Therefore, in this paper, we make some initial progress towards reintroducing behavior tokens in LLM training. The trained models, other than showing similar performance to LLMs on content understanding tasks, show generalization capabilities on behavior simulation, content simulation, behavior understanding, and behavior domain adaptation. Using a wide range of tasks on two corpora, we show results on all these capabilities. We call these models Large Content and Behavior Models (LCBMs). Further, to spur more research on LCBMs, we release our new Content Behavior Corpus (CBC), a repository containing communicator, message, and corresponding receiver behavior.

Coding agents and LLM-powered applications routinely send potentially sensitive content to cloud LLM APIs where it may be logged, retained, used for training, or subpoenaed. Existing privacy tooling focuses on network-level encryption and organization-level DLP, neither of which addresses the content of prompts themselves. We present a systematic empirical evaluation of eight techniques for privacy-preserving LLM requests: (A) local-only inference, (B) redaction with placeholder restoration, (C) semantic rephrasing, (D) Trusted Execution Environment hosted inference, (E) split inference, (F) fully homomorphic encryption, (G) secret sharing via multi-party computation, and (H) differential-privacy noise. We implement all eight (or a tractable research-stage subset where deployment is not yet feasible) in an open-source shim compatible with MCP and any OpenAI-compatible API. We evaluate the four practical options (A, B, C, H) and their combinations across four workload classes using a ground-truth-labelled leak benchmark of 1,300 samples with 4,014 annotations. Our headline finding is that no single technique dominates: the combination A+B+C (route locally when possible, redact and rephrase the rest) achieves 0.6% combined leak on PII and 31.3% on proprietary code, with zero exact leaks on PII across 500 samples. We present a decision rule that selects the appropriate option(s) from a threat-model budget and workload characterisation. Code, benchmarks, and evaluation harness are released at https://github.com/jayluxferro/llm-redactor.

Advancements in large language models (LLMs) have unlocked remarkable capabilities. While deploying these models typically requires server-grade GPUs and cloud-based inference, the recent emergence of smaller open-source models and increasingly powerful consumer devices have made on-device deployment practical. The web browser as a platform for on-device deployment is universally accessible, provides a natural agentic environment, and conveniently abstracts out the different backends from diverse device vendors. To address this opportunity, we introduce WebLLM, an open-source JavaScript framework that enables high-performance LLM inference entirely within web browsers. WebLLM provides an OpenAI-style API for seamless integration into web applications, and leverages WebGPU for efficient local GPU acceleration and WebAssembly for performant CPU computation. With machine learning compilers MLC-LLM and Apache TVM, WebLLM leverages optimized WebGPU kernels, overcoming the absence of performant WebGPU kernel libraries. Evaluations show that WebLLM can retain up to 80% native performance on the same device, with room to further close the gap. WebLLM paves the way for universally accessible, privacy-preserving, personalized, and locally powered LLM applications in web browsers. The code is available at: https://github.com/mlc-ai/web-llm.

We present LLaDA2.0-Uni, a unified discrete diffusion large language model (dLLM) that supports multimodal understanding and generation within a natively integrated framework. Its architecture combines a fully semantic discrete tokenizer, a MoE-based dLLM backbone, and a diffusion decoder. By discretizing continuous visual inputs via SigLIP-VQ, the model enables block-level masked diffusion for both text and vision inputs within the backbone, while the decoder reconstructs visual tokens into high-fidelity images. Inference efficiency is enhanced beyond parallel decoding through prefix-aware optimizations in the backbone and few-step distillation in the decoder. Supported by carefully curated large-scale data and a tailored multi-stage training pipeline, LLaDA2.0-Uni matches specialized VLMs in multimodal understanding while delivering strong performance in image generation and editing. Its native support for interleaved generation and reasoning establishes a promising and scalable paradigm for next-generation unified foundation models. Codes and models are available at https://github.com/inclusionAI/LLaDA2.0-Uni.

The rise of LLMs has driven demand for private serverless deployments, characterized by moderate-scale models and infrequent requests. While existing solutions follow exclusive GPU deployment, we take a step back to explore modern platforms and find that: Emerging CPU architectures with built-in accelerators are capable of serving LLMs but remain underutilized, and both CPUs and GPUs can accommodate multiple LLMs simultaneously. We propose LLM-Mesh, a serverless inference scheme for small-to-mid-sized LLMs that enables elastic sharing across heterogeneous hardware. LLM-Mesh tackles three fundamental challenges: (1) precise, fine-grained compute resource allocation at token-level to handle fluctuating computational demands; (2) a coordinated and forward-looking memory scaling mechanism to detect out-of-memory hazards and reduce operational overhead; and (3) a dual approach that reduces resource fragmentation through proactive preemption and reactive bin-packing. Experimental results on 4 32-core CPUs and 4 A100 GPUs show that LLM-Meshimproves service capacity by 44% - 63% through sharing, while further leveraging CPUs boosts this to 91% - 159%.

Large language models (LLMs) have enabled the creation of multi-modal LLMs that exhibit strong comprehension of visual data such as images and videos. However, these models usually rely on extensive visual tokens from visual encoders, leading to high computational demands, which limits their applicability in resource-constrained environments and for long-context tasks. In this work, we propose a training-free adaptive inference method for multi-modal LLMs that can accommodate a broad range of efficiency requirements with a minimum performance drop. Our method consists of a) iterative token merging based on embedding similarity before LLMs, and b) progressive token pruning within LLM layers based on multi-modal importance. With a minimalist design, our method can be applied to both video and image LLMs. Extensive experiments on diverse video and image benchmarks demonstrate that, our method substantially reduces computation load (e.g., a 7-fold reduction in FLOPs) while preserving the performance of video and image LLMs. Further, under a similar computational cost, our method outperforms the state-of-the-art methods in long video understanding (e.g., +4.6 on MLVU). Additionally, our in-depth analysis provides insights into token redundancy and LLM layer behaviors, offering guidance for future research in designing efficient multi-modal LLMs. Our code will be available at https://github.com/LaVi-Lab/AIM.