Codestral on Apple M3 Max · 36GB

✅ Runs (tight) — Codestral 22B @ Q5_K_M
Codestral 22B at Q5_K_M needs ~17.2 GB (weights 14.7 GB + KV 1.8 GB + overhead 755 MB @ 8K ctx) of your 21.1 GB usable (of 36.0 GB unified memory) — fits, but little headroom — close other apps or trim context. Expect ~10–17 tok/s (usable).
context @ Q5_K_M: comfortable to 4K · runs to its full 32K
quantneedsspeed
~ Q8_024.9 GB~0.9–1 tok/s (slow)
Q5_K_M17.2 GB~10–17 tok/s (usable)
Q4_K_M14.9 GB~12–20 tok/s (usable)
Q3_K_M12.4 GB~15–25 tok/s (usable)
$runlocal install codestral:22b
runlocal verdict card
download card share on x share on reddit drop it in your model-drop post
Codestral: Runs (tight)
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 M3 Max: Qwen3 · DeepSeek-R1 (distill) · Qwen2.5 · Gemma 4 · Gemma 3

Best models for a Apple M3 Max · 36GB →

Architecturally similar

These share Codestral's KV-cache geometry — they size memory the same way, so here's how they fit on a Apple M3 Max · 36GB too:

modelsharesverdict here
Llama 3.1 8kv/128hd ✅ Runs comfortably
Qwen2.5-Coder 8kv/128hd ✅ Runs (tight)
Llama 3.2 8kv/128hd ✅ Runs comfortably
Llama 3.3 8kv/128hd 🐢 CPU-only (slow)
Mistral 7B 8kv/128hd ✅ Runs comfortably
Mistral Small 3 8kv/128hd ✅ Runs (tight)
Mistral Nemo 8kv/128hd ✅ Runs comfortably
Llama 3.2 Vision 8kv/128hd ✅ Runs comfortably

…and 8 more — see the Codestral page.

Check any combo yourself: open the checker →