MLX Community · First MLX build 2 Mar 2026
Qwen3.5-0.8B
MLX
TextVision apache-2.0
A 853M parameter vision and text model, with 16 MLX builds, from 3-bit to bf16.
- Parameters
- 853M From Hugging Face
- Quantisation
- 4-bit Qwen3.5-0.8B-4bit
- Memory needed
- 3 GB Estimate, unified memory
- Download
- 625 MB Weights, estimated
- Downloads
- 60K This build, last 30 days
- Task
- Vision and text Text, Vision
- MLX builds
- 16 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 3 GB at 4-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 · mlx-vlm$ pip install -U mlx-vlm
$ python -m mlx_vlm.generate --model mlx-community/Qwen3.5-0.8B-4bit --image photo.jpg --prompt "Describe this image"MLX builds
16 on mlx-community| Qwen3.5-0.8B-3bit | 3-bit | 280 MB | 3 GB | 29 |
|---|---|---|---|---|
| Qwen3.5-0.8B-mxfp4 | MXFP4 | 281 MB | 3 GB | 36 |
| Qwen3.5-0.8B-mixed_3_4 | 4-bit | 294 MB | 3 GB | 19 |
| Qwen3.5-0.8B-mixed_2_6 | 4-bit | 297 MB | 3 GB | 21 |
| Qwen3.5-0.8B-mixed_3_5 | 4-bit | 299 MB | 3 GB | 16 |
| Qwen3.5-0.8B-mixed_3_6 | 4-bit | 305 MB | 3 GB | 21 |
| Qwen3.5-0.8B-nvfp4 | 4-bit | 308 MB | 3 GB | 1.5K |
| Qwen3.5-0.8B-mixed_4_6 | 4-bit | 313 MB | 3 GB | 28 |
| Qwen3.5-0.8B-mixed_3_8 | 4-bit | 315 MB | 3 GB | 12 |
| Qwen3.5-0.8B-mixed_4_8 | 4-bit | 323 MB | 3 GB | 18 |
| Qwen3.5-0.8B-mxfp8 | MXFP8 | 427 MB | 3 GB | 761 |
| Qwen3.5-0.8B-4bit | 4-bit | 625 MB | 3 GB | 60K |
| Qwen3.5-0.8B-5bit | 5-bit | 719 MB | 3 GB | 586 |
| Qwen3.5-0.8B-6bit | 6-bit | 813 MB | 3 GB | 338 |
| Qwen3.5-0.8B-8bit | 8-bit | 1.0 GB | 4 GB | 7.3K |
| Qwen3.5-0.8B-bf16 | bf16 | 1.7 GB | 4 GB | 2.0K |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.