Qwen3 Coder 30B-A3B on NVIDIA GeForce RTX 4090 · 24GB VRAM

✅ Runs (tight) — Qwen3 Coder 30B-A3B @ Q4_K_M
Qwen3 Coder 30B-A3B 30B-A3B (MoE) at Q4_K_M needs ~19.0 GB (weights 17.4 GB + KV 768 MB + overhead 888 MB @ 8K ctx) of your 21.6 GB usable VRAM — fits, but little headroom — close other apps or trim context. Expect ~150–250 tok/s (very fast).
context @ Q4_K_M: runs to 128K · won't fit past that on this rig
quantneedsspeed
✗ Q8_032.5 GB—
~ Q6_K25.3 GB~35–60 tok/s (fast)
✓ Q4_K_M19.0 GB~150–250 tok/s (very fast)
$runlocal install qwen3-coder:30b
⚠ NVIDIA support is best-effort in v0.1 — verify before relying on it.
runlocal verdict card
download card share on x share on reddit drop it in your model-drop post
Qwen3 Coder 30B-A3B: 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 NVIDIA GeForce RTX 4090: DeepSeek-R1 (distill) · Qwen3 · Qwen2.5 · Gemma 3 · Gemma 4

Best models for a NVIDIA GeForce RTX 4090 · 24GB VRAM →

Architecturally similar

These share Qwen3 Coder 30B-A3B's KV-cache geometry — they size memory the same way, so here's how they fit on a NVIDIA GeForce RTX 4090 · 24GB VRAM too:

modelsharesverdict here
Qwen2.5-Coder 4kv/128hd ✅ Runs (tight)
StarCoder2 4kv/128hd ✅ Runs comfortably
Yi 1.5 4kv/128hd ✅ Runs (tight)
SmolLM3 3B 4kv/128hd ✅ Runs comfortably
Yi-Coder 9B 4kv/128hd ✅ Runs comfortably

Check any combo yourself: open the checker →