🔌
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 M1 Ultra 128GB run Qwen2.5 Instruct (32B)?
Yes — here's how fast and up to what context, with real measured numbers.
Qwen2.5 Instruct (32B) on the Apple M1 Ultra 128GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 186k | 8.4-10.1 t/s | ~0.1-0.1 (slow) |
| Q8 | yes | 310k | 16.7-20.2 t/s | ~0.2-0.3 (slow) |
| Q4 | yes | 364k | 29.9-36 t/s | ~0.4-0.5 (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 M1 Ultra 128GB run Qwen2.5 Instruct (32B)?▶
Yes. At Q4 it generates ~29.9-36 tok/s and fits up to 364k context; beyond that it spills to SSD swap and slows to ~0.4-0.5 tok/s.
How much context fits?▶
At Q4, up to about 364k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.