FunAudioLLM · First MLX build 2 Jul 2025
CosyVoice2-0.5B
MLX
Audio apache-2.0
A 782M parameter text to speech model from FunAudioLLM, with 4 MLX builds, from 4-bit to fp16.
- Parameters
- 782M From Hugging Face
- Quantisation
- 8-bit CosyVoice2-0.5B-8bit
- Memory needed
- 4 GB Estimate, unified memory
- Download
- 957 MB Weights, estimated
- Downloads
- 802 This build, last 30 days
- Task
- Text to speech Audio
- MLX builds
- 4 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 4 GB at 8-bit
8 GB
16 GB
24 GB
32 GB
64 GB
128 GB
Fits with headroomTight, close other appsWon’t fit
Estimated from the size of the weights, plus a fifth for context and working memory and 2 GB for macOS and your apps. Long contexts need more.
Run it
Terminal · Hugging Face CLI$ pip install -U huggingface_hub
$ hf download mlx-community/CosyVoice2-0.5B-8bitMLX builds
4 on mlx-community| CosyVoice2-0.5B-4bit | 4-bit | 778 MB | 3 GB | 85 |
|---|---|---|---|---|
| CosyVoice2-0.5B-8bit | 8-bit | 957 MB | 4 GB | 802 |
| CosyVoice2-0.5B-fp16 | fp16 | 1.6 GB | 4 GB | 115 |
| CosyVoice2-0.5B-S3Tokenizer | Unstated | — | — | 0 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.