🔌
New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
// RUN — FITS
Can the Apple M3 Max 36GB run Qwen-VLA?
Yes — here's how fast and up to what context, with real measured numbers.
Qwen-VLA on the Apple M3 Max 36GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 157k | 19.8-23.9 t/s | ~0.7-0.8 (slow) |
| Q8 | yes | 196k | 39.6-47.8 t/s | ~1.3-1.6 (slow) |
| Q4 | yes | 214k | 70.7-85.3 t/s | ~2.4-2.8 (slow) |
Estimate (memory-bound). Beyond "max context" the KV cache spills to SSD swap and speed drops to the "beyond" figure.
FAQ
Can the Apple M3 Max 36GB run Qwen-VLA?▶
Yes. At Q4 it generates ~70.7-85.3 tok/s and fits up to 214k context; beyond that it spills to SSD swap and slows to ~2.4-2.8 tok/s.
How much context fits?▶
At Q4, up to about 214k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.