llama.cpp on the M4 Max 64GB
llamaperf has 6 llama.cpp reports on the M4 Max 64GB, none of them a plain single-card run with a model size and quant, so there is no median to quote. The reports are listed below with how each one ran.
6 llama.cpp reports on the M4 Max 64GB
Newest first. A run on several cards, with speculative decoding, with part of the model in system RAM or with batched requests says so, since each describes a different setup.
- Qwen3.8 · Q4_K_M23.9 t/s
- Qwen3.6 · UD-Q4_K_XL76.4 t/s
- Qwen3.5 · Q4_K_M65.5 t/s
- Gemma 4 · Q4_K_M22.4 t/s
- Gemma 4 · UD-Q4_K_M81.7 t/s
- Qwen3.8 125B · 6B active · AD-3.84bpw-M6436.0 t/s(part in system RAM)
Frequently asked
How fast is llama.cpp on the M4 Max 64GB?
llamaperf has 6 llama.cpp reports on the M4 Max 64GB, none of them a plain single-card run with a model size and quant, so there is no median to quote. The reports are listed below with how each one ran.