🔌
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 M4 Max 48GB run EmbeddingGemma?
Yes — here's how fast and up to what context, with real measured numbers.
EmbeddingGemma on the Apple M4 Max 48GB
| Quant | Fits? | Max context | Speed | Beyond context |
|---|---|---|---|---|
| FP16 | yes | 627k | 602.7-726.8 t/s | ~11-13.3 (slow) |
| Q8 | yes | 632k | 1205.5-1453.6 t/s | ~22.1-26.6 (slow) |
| Q4 | yes | 634k | 2152.6-2595.8 t/s | ~39.4-47.5 (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 M4 Max 48GB run EmbeddingGemma?▶
Yes. At Q4 it generates ~2152.6-2595.8 tok/s and fits up to 634k context; beyond that it spills to SSD swap and slows to ~39.4-47.5 tok/s.
How much context fits?▶
At Q4, up to about 634k tokens stay in fast memory; longer context spills to SSD swap and throughput collapses.