Mistral Small 3 on Apple M4 Max · 128GB

✅ Runs comfortably — Mistral Small 3 24B @ Q8_0
Mistral Small 3 24B at Q8_0 needs ~25.8 GB (weights 23.3 GB + KV 1.3 GB + overhead 1.2 GB @ 8K ctx) of your 93.0 GB usable (of 128 GB unified memory) — plenty of headroom. Expect ~9–15 tok/s (usable).
context @ Q8_0: comfortable to 32K
quantneedsspeed
✓ Q8_025.8 GB~9–15 tok/s (usable)
✓ Q5_K_M17.6 GB~13–20 tok/s (usable)
✓ Q4_K_M15.3 GB~15–25 tok/s (fast)
✓ Q3_K_M12.5 GB~19–30 tok/s (fast)
$runlocal install mistral-small:24b
runlocal verdict card
download card share on x share on reddit drop it in your model-drop post
Mistral Small 3: Runs comfortably
for your own README — links back here

Same model, other machines: Apple M1 · Apple M1 Pro · Apple M2 · Apple M3 Pro · Apple M2 Pro · Apple M4 Pro

Also runs on a Apple M4 Max: DeepSeek-R1 (distill) · Qwen3 · Qwen2.5 · Llama 3.1 · Llama 3.3

Best models for a Apple M4 Max · 128GB →

Architecturally similar

These share Mistral Small 3's KV-cache geometry — they size memory the same way, so here's how they fit on a Apple M4 Max · 128GB too:

modelsharesverdict here
Qwen2.5-Coder 8kv/128hd ✅ Runs comfortably
Llama 3.2 8kv/128hd ✅ Runs comfortably
Mistral 7B 8kv/128hd ✅ Runs comfortably
Mistral Nemo 8kv/128hd ✅ Runs comfortably
Codestral 8kv/128hd ✅ Runs comfortably
Llama 3.2 Vision 8kv/128hd ✅ Runs comfortably
Mixtral 8x7B 8kv/128hd ✅ Runs comfortably
Phi-4-mini 8kv/128hd ✅ Runs comfortably

…and 6 more — see the Mistral Small 3 page.

Check any combo yourself: open the checker →