community-verified · measured pool data

Can the M3 Ultra 512GB run Mistral-Small-3.2-24B-Instruct-2506?

Yes — community-measured at 39.5 tok/s (2 runs at 4-bit).

Verified single-stream runs on the M3 Ultra 512GB from the localmaxxing public pool. Every number on this page is measured — nothing here is estimated.

Quantization	Median tok/s	TTFT (ms)	Peak VRAM	Max context tested	Engines
4-bit	39.5	129	—	8,192	llama.cpp

Go deeper

I have a goal — find models that fit my use case → I have hardware — browse everything my rig can run → All machine × model answer pages →