--- license: mit pipeline_tag: text-to-speech library_name: rlx tags: - chatterbox - tts - onnx - rlx --- # ChatterBox ONNX + RLX native (staging) ChatterBox ONNX tree plus RLX `native/` T3 extract (parity bins removed). | Field | Value | |---|---| | **Hub id** | [`eugenehp/chatterbox`](https://huggingface.co/eugenehp/chatterbox) | | **Kind** | Converted / re-laid-out for RLX from an upstream checkpoint. | | **RLX crate** | [`rlx-chatterbox`](https://github.com/MIT-RLX/rlx-models/tree/main/crates/rlx-chatterbox) | | **Upstream** | https://huggingface.co/synath/chatterbox-ONNX | ## Quick start ```bash hf download eugenehp/chatterbox --local-dir . cargo run -p rlx-chatterbox --release -- --model-dir . ``` ## File highlights - `native/t3_lm.safetensors` (976.2 MiB) - `t3-q4_k_m.gguf` (299.6 MiB) - `onnx/cfm_estimator.onnx` (273.0 MiB) - `onnx/flow_encoder.onnx` (177.2 MiB) - `onnx/hift_head.onnx` (79.6 MiB) - `onnx/hift_stages.onnx` (79.6 MiB) - `weights.safetensors` (77.6 MiB) - `native/embed_tables.safetensors` (58.8 MiB) - `native/rope_cache.safetensors` (32.0 MiB) - `onnx/hift_src.onnx` (12.6 MiB) - `onnx/hift_f0.onnx` (12.5 MiB) - `onnx/conditional_decoder.onnx` (6.1 MiB) - `onnx/speech_encoder.onnx` (1.1 MiB) - `default_voice.wav` (697.6 KiB) - `onnx/language_model_q4f16.onnx` (224.0 KiB) - `onnx/language_model_fp16.onnx` (168.6 KiB) ## Run with RLX Clone [rlx-models](https://github.com/MIT-RLX/rlx-models), place this repo under `weights/tts/chatterbox` (or pass the path explicitly), then: ```bash cargo run -p rlx-chatterbox --release -- --model-dir . ``` ## License MIT — see `LICENSE`. Inherit upstream terms when redistributing. Original weights and authorship: https://huggingface.co/synath/chatterbox-ONNX ## Maintenance Cards and LFS attrs are regenerated from the local `weights/` tree in [rlx-models](https://github.com/MIT-RLX/rlx-models) via `python3 scripts/prepare_weights_hf.py`.