MLX Community · First MLX build 18 Apr 2026
higgs-audio-v2-3B
MLX
Audio apache-2.0
A text to speech model, with 2 MLX builds, from 8-bit to 6-bit.
- Parameters
- — Not published
- Quantisation
- 8-bit higgs-audio-v2-3B-mlx-q8
- Memory needed
- — GB Estimate, unified memory
- Download
- — Weights, estimated
- Downloads
- 600 This build, last 30 days
- Task
- Text to speech Audio
- MLX builds
- 2 On mlx-community
- Devices
- Mac Apple silicon
Memory
Hugging Face does not list a weights size for this build, so there is no memory estimate.
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/higgs-audio-v2-3B-mlx-q8MLX builds
2 on mlx-community| higgs-audio-v2-3B-mlx-q8 | 8-bit | — | — | 600 |
|---|---|---|---|---|
| higgs-audio-v2-3B-mlx-q6 | 6-bit | — | — | 136 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.