🔌
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 48GB run FRED-T5-XL?
Yes — here's how fast and up to what context, with real measured numbers.
FRED-T5-XL on the Apple M3 Max 48GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 583k | 78.2-94.3 t/s | ~2-2.4 (slow) |
| Q8 | yes | 610k | 156.3-188.5 t/s | ~3.9-4.7 (slow) |
| Q4 | yes | 621k | 279.1-336.6 t/s | ~7-8.4 (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 48GB run FRED-T5-XL?▶
Yes. At Q4 it generates ~279.1-336.6 tok/s and fits up to 621k context; beyond that it spills to SSD swap and slows to ~7-8.4 tok/s.
How much context fits?▶
At Q4, up to about 621k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.