Qwen3 Coder 30B-A3B Instruct › Apple M2 Ultra
Qwen3 Coder 30B-A3B Instruct on the Apple M2 Ultra
Best measured: 79.0 tokens/s generation and 2,453 tokens/s prompt processing (Q8_0, Metal, FA on).
For comparison, our size-scaling estimate for this combination is 99.3 tok/s.
Every run
| Quant | Backend | Generate tok/s | Prompt tok/s | Test | Build / by / notes | Source |
|---|---|---|---|---|---|---|
| Q8_0 | Metal FA on | 79.0 | 2,453 | tg32 | ggml-org maintainers Benchmarked repo: https://huggingface.co/ggml-org/Qwen3-Coder-30B-A3B-Instruct-Q8_0-GGUF; chip named only by the file's directory (M2 Ultra), GPU core count not stated | llama.cpp benches/ (maintainer runs) |
tg128 = generating 128 tokens; pp512 = reading a 512-token prompt (llama-bench). LocalScore averages several prompt and output lengths, so its numbers run lower than tg128.
No card yet?
Try it on a rented GPU (RunPod)Affiliate link: if you sign up, RigBench may earn a referral credit. It never affects rankings or numbers.
All results for the Apple M2 Ultra · Qwen3 Coder 30B-A3B Instruct on other hardware