Qwen3 Coder 30B-A3B on NVIDIA GeForce RTX 5090 · 32GB VRAM

✅ Runs (tight) — Qwen3 Coder 30B-A3B @ Q6_K
Qwen3 Coder 30B-A3B 30B-A3B (MoE) at Q6_K needs ~25.3 GB (weights 23.4 GB + KV 768 MB + overhead 1.2 GB @ 8K ctx) of your 28.8 GB usable VRAM — fits, but little headroom — close other apps or trim context. Expect ~210–350 tok/s (very fast).
context @ Q6_K: runs to 128K · won't fit past that on this rig
quantneedsspeed
~ Q8_032.5 GB~55–90 tok/s (very fast)
✓ Q6_K25.3 GB~210–350 tok/s (very fast)
✓ Q4_K_M19.0 GB~280–460 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 5090: DeepSeek-R1 (distill) · Qwen2.5 · Qwen3 · Llama 3.1 · Gemma 3

Best models for a NVIDIA GeForce RTX 5090 · 32GB 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 5090 · 32GB 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 →