Qwen3.6 35B A3B has 35 billion parameters in total, of which 3 billion are active for each token, so it writes quickly for its size. Qwen highlights agentic coding and an option to keep reasoning from earlier turns. It reads text, images and video and has a 262,144-token context, extensible to about 1 million.
A 4-bit MLX build is about 20 GB; on mlx.app's estimate it needs a Mac with 32 GB of memory.
Open weights under Apache 2.0. Run locally, nothing leaves your Mac.
How to Get It
Official downloads only. MLX never re-hosts an app, and never runs a command for you.
MLX Score
Ranked against the open models listings, from real public data only. How the score works.
Information
- Developer
- Qwen (Alibaba)
- Category
- Open models
- Price
- Free (Open weights, free to download.)
- Your data
- Runs on your Mac
- Licence
- Apache-2.0
- Mac
- macOS 14.0 or later
- iPhone and iPad
- —
- Needs
- 32 GB memory
- Links
- Website · Hugging Face
- Facts checked
- 2026-11-01
Sources
You Might Also Like
Gemma 4 31BModels
gpt-oss-20bModels
Llama 3.1 8B InstructModels
Qwen3.5 9BModels
DeepSeek V4 Flash 0731Models
gpt-oss-120bModels
Gemma 4 E4BModels
Qwen3.6 27BModels
Nemotron 3 Nano 30B A3BModels
GLM-4.7 FlashModels
Qwen3 Coder 30B A3B InstructModels
Llama 3.3 70B InstructModelsQwen3.6 35B A3B is a product of Qwen (Alibaba); MLX isn’t affiliated with it. Something wrong? Write to support@mlx.app.