Can the Apple M3 Ultra run Qwen3 235B-A22B?
Yes, with 256 GB or more. Qwen3 235B-A22B needs about 144 GB and the Apple M3 Ultra with 256 GB of unified memory leaves about 192 GB for models. Expect about 22 tok/s, about reading speed. The 96 GB version doesn't have room.
ollama run qwen3:235bFirst time? Install Ollama, then paste that into a terminal.
Each Apple M3 Ultra version
- 96 GB (about 72 GB usable)Too big
- 256 GB (about 192 GB usable)Usable ~22 tok/s
- 512 GB (about 384 GB usable)Usable ~22 tok/s
Speed comes from memory bandwidth: the card reads all 13.3 GB of active weights for every token. On the Apple M3 Ultra, gpt-oss 120B runs at about 99 tok/s if you can trade some quality for speed. Cards that run Qwen3 235B-A22B faster:
No single card or Mac we track does. It needs a multi-GPU rig; see the builds.
Speed is roughly memory bandwidth divided by the weights read per token, calibrated on our own rack. Sizes are Ollama's default 4-bit downloads plus about 1.2 GB for context. Real results are usually within about a third. Measured it yourself? Send us the number.