🔌
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 M2 Max 96GB run TRIMELMlong (150M)?
Yes — here's how fast and up to what context, with real measured numbers.
TRIMELMlong (150M) on the Apple M2 Max 96GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 1291k | 906.7-1093.3 t/s | ~22.7-27.3 (slow) |
| Q8 | yes | 1293k | 1813.3-2186.7 t/s | ~45.3-54.7 (slow) |
| Q4 | yes | 1294k | 3238.1-3904.8 t/s | ~81-97.6 (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 M2 Max 96GB run TRIMELMlong (150M)?▶
Yes. At Q4 it generates ~3238.1-3904.8 tok/s and fits up to 1294k context; beyond that it spills to SSD swap and slows to ~81-97.6 tok/s.
How much context fits?▶
At Q4, up to about 1294k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.