---
language:
- en
- zh
- hi
- ta
library_name: transformers
license: mit
pipeline_tag: text-generation
datasets:
- McAuley-Lab/Reddit2Deezer
metrics:
- exact_match
base_model:
-  fluidity-vapour-4.1
tags:
- best
- code
---

# fluidity-vapour-4.2


## Introduction

We're introducing fluidity-vapour-4.1, our latest flagship model for long-horizon tasks. It marks a substantial leap in long-horizon task capability over its predecessor fluidity-vapour-4.1 and, for the first time, delivers that capability on a **solid 1M-token context**. fluidity-vapour-4.2 new capabilities include:
- **Solid 1M Context:** A solid 1M-token context that stably sustains long-horizon work
- **Advanced Coding with Flexible Effort**: Stronger coding capabilities with multiple thinking effort levels to balance performance and latency
- **Improved Architecture**: We propose [IndexShare](https://arxiv.org/abs/2603.12201), which reuses the same indexer across every four sparse attention layers, reducing per-token FLOPs by 2.9× at a 1M context length. We also improve GLM-5.2’s MTP layer for speculative decoding, increasing the acceptance length by up to 20%
- **Pure Open**: An MIT open-source license — no regional limits, technical access without borders

https://raw.githubusercontent.com/FluidityVapour/FV-4.2/refs/heads/main/resources/water_drop.png

## Benchmark
BenchmarkFluidity-Vapour 4.2GLM-5.1Qwen3.7-MaxMiniMax M3DeepSeek-V4-ProClaude Opus 4.8GPT-5.5Gemini 3.1 ProReasoningHLE40.53141.43737.749.8*41.4*45HLE (w/ Tools)54.752.353.5-48.257.9*52.2*51.4*CritPt20.94.613.43.712.920.927.117.7AIME 202699.295.397-94.695.798.398.2HMMT Nov. 202594.4949584.494.496.596.594.8HMMT Feb. 202692.582.697.184.495.296.796.787.3IMOAnswerBench91.083.890-89.883.5-81GPQA-Diamond91.286.2909390.193.693.694.3CodingSWE-bench Pro62.158.460.65955.469.258.654.2NL2Repo48.942.747.242.135.569.750.733.4DeepSWE46.21818208587010ProgramBench63.750.9--47.871.970.839.5Terminal Bench 2.1 (Terminus-2)81.063.5756564858474Terminal Bench 2.1 (Best Reported Harness)82.769---78.983.470.7FrontierSWE (Dominance)74.430.5--29.075.172.639.6PostTrainBench34.320.1---37.228.421.6SWE-marathon13.01.0---26.012.04.0AgenticMCP-Atlas (Public Set)76.871.876.474.273.677.875.369.2Tool-Decathlon48.240.7--52.859.955.648.8

## Serve fluidity-vapour-4.2 Locally

fluidity-vapour-4.2 supports deployment with the following frameworks. Feel free to try them out:

- [SGLang](https://github.com/sgl-project/sglang) (v0.5.13.post1+) — see [cookbook](https://cookbook.sglang.io/autoregressive/GLM/GLM-5.2)
- [vLLM](https://github.com/vllm-project/vllm) (v0.23.0+) — see [recipes](https://recipes.vllm.ai/zai-org/GLM-5.2)
- [Transformers](https://github.com/huggingface/transformers) (v0.5.12+) — see [transformers docs](https://github.com/huggingface/transformers/blob/main/docs/source/en/model_doc/glm_moe_dsa.md)
- [KTransformers](https://github.com/kvcache-ai/ktransformers) (v0.5.12+) — see [tutorial](https://github.com/kvcache-ai/ktransformers/blob/main/doc/en/kt-kernel/GLM-5.2-Tutorial.md)
- [Unsloth](https://github.com/unslothai/unsloth) (v0.1.47-beta+) — see [guide](https://unsloth.ai/docs/models/glm-5.2)
- For deployment on the `Ascend NPU` platform, inference frameworks such as vLLM-Ascend, xLLM and SGLang are supported — see [here](github.com/zai-org/GLM-5/blob/main/example/ascend.md).

## Citation

If you find fluidity-vapour-4.2 useful in your research, please cite our technical report: