MLX Community · First MLX build 31 Mar 2026
LongCat-AudioDiT-3.5B
MLX
Audio
A 3.8B parameter text to speech model, with 8 MLX builds, from 4-bit to bf16.
- Parameters
- 3.8B From Hugging Face
- Quantisation
- 8-bit LongCat-AudioDiT-3.5B-8bit
- Memory needed
- 8 GB Estimate, unified memory
- Download
- 4.5 GB Weights, estimated
- Downloads
- 85 This build, last 30 days
- Task
- Text to speech Audio
- MLX builds
- 8 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 8 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/LongCat-AudioDiT-3.5B-8bitMLX builds
8 on mlx-community| LongCat-AudioDiT-3.5B-4bit | 4-bit | 2.7 GB | 6 GB | 25 |
|---|---|---|---|---|
| LongCat-AudioDiT-3.5B-mxfp4 | MXFP4 | 2.7 GB | 6 GB | 15 |
| LongCat-AudioDiT-3.5B-nvfp4 | 4-bit | 2.7 GB | 6 GB | 10 |
| LongCat-AudioDiT-3.5B-5bit | 5-bit | 3.2 GB | 6 GB | 16 |
| LongCat-AudioDiT-3.5B-6bit | 6-bit | 3.6 GB | 7 GB | 17 |
| LongCat-AudioDiT-3.5B-8bit | 8-bit | 4.5 GB | 8 GB | 85 |
| LongCat-AudioDiT-3.5B-mxfp8 | MXFP8 | 4.5 GB | 8 GB | 12 |
| LongCat-AudioDiT-3.5B-bf16 | bf16 | 15.3 GB | 20 GB | 78 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.