Gemma 3 on Apple M3 Max · 36GB

✅ Runs (tight) — Gemma 3 27B @ Q4_K_M
Gemma 3 27B at Q4_K_M needs ~20.1 GB (weights 15.5 GB + KV 3.9 GB + overhead 793 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 ~9–14 tok/s (usable).
context @ Q4_K_M: runs to 32K · won't fit past that on this rig
quantneedsspeed
~ Q8_032.4 GB~0.7–1 tok/s (slow)
~ Q5_K_M23.0 GB~1–2 tok/s (slow)
✓ Q4_K_M20.1 GB~9–14 tok/s (usable)
✓ Q3_K_M17.0 GB~10–17 tok/s (usable)
$runlocal install gemma3:27b
runlocal verdict card
download card share on x share on reddit drop it in your model-drop post
Gemma 3: 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 · Mistral Small 3

Best models for a Apple M3 Max · 36GB →

Architecturally similar

These share Gemma 3'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
Gemma 2 16kv/128hd4kv/256hd8kv/256hd ✅ Runs (tight)
Falcon3 7B 4kv/256hd ✅ Runs comfortably

Check any combo yourself: open the checker →