MLX Community · First MLX build 21 Feb 2026
Ming-omni-tts-16.8B-A3B
MLX
Audio apache-2.0
A 18B parameter text to speech model, with 5 MLX builds, from 4-bit to 8-bit.
- Parameters
- 18B From Hugging Face
- Quantisation
- bf16 Ming-omni-tts-16.8B-A3B-bf16
- Memory needed
- 42 GB Estimate, unified memory
- Download
- 35.8 GB Weights, estimated
- Downloads
- 61 This build, last 30 days
- Task
- Text to speech Audio
- MLX builds
- 5 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 42 GB at bf16
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/Ming-omni-tts-16.8B-A3B-bf16MLX builds
5 on mlx-community| Ming-omni-tts-16.8B-A3B-4bit | 4-bit | 2.4 GB | 5 GB | 52 |
|---|---|---|---|---|
| Ming-omni-tts-16.8B-A3B-bf16 | bf16 | 35.8 GB | 42 GB | 61 |
| Ming-omni-tts-16.8B-A3B-5bit | 5-bit | — | — | 0 |
| Ming-omni-tts-16.8B-A3B-6bit | 6-bit | — | — | 0 |
| Ming-omni-tts-16.8B-A3B-8bit | 8-bit | — | — | 0 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.