🔌
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 phi-3.5-mini?
Yes — here's how fast and up to what context, with real measured numbers.
phi-3.5-mini on the Apple M3 Max 36GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 356k | 26.8-32.4 t/s | ~0.9-1.1 (slow) |
| Q8 | yes | 414k | 53.7-64.7 t/s | ~1.8-2.2 (slow) |
| Q4 | yes | 439k | 95.9-115.6 t/s | ~3.2-3.9 (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 phi-3.5-mini?▶
Yes. At Q4 it generates ~95.9-115.6 tok/s and fits up to 439k context; beyond that it spills to SSD swap and slows to ~3.2-3.9 tok/s.
How much context fits?▶
At Q4, up to about 439k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.