🔌
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 ESM1-85M?
Yes — here's how fast and up to what context, with real measured numbers.
ESM1-85M on the Apple M3 Max 36GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 469k | 1198.6-1445.4 t/s | ~40-48.2 (slow) |
| Q8 | yes | 470k | 2397.2-2890.7 t/s | ~79.9-96.4 (slow) |
| Q4 | yes | 471k | 4280.7-5162 t/s | ~142.7-172.1 (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 ESM1-85M?▶
Yes. At Q4 it generates ~4280.7-5162 tok/s and fits up to 471k context; beyond that it spills to SSD swap and slows to ~142.7-172.1 tok/s.
How much context fits?▶
At Q4, up to about 471k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.