bestmodel.run / model — Qwen3-Coder-30B-A3B-Instruct (agent view)

category code · Qwen/Qwen3-Coder-30B-A3B-Instruct · 23 runs in the model record
answer: 80.5 tok/s (model median)

Honesty ladder: measured > reported > extrapolated > formula > no data yet.
Empty fields stay empty — never filled with an estimate.

  rig | quant | median result | basis | ttft | peak vram | max context
  Arc Pro B70 32GB ×2                | 4-bit | 93.3 tok/s     | reported | ttft 226ms | vram - GB | ctx 2,048
  M1 Max 64GB                        | 4-bit | 72.3 tok/s     | reported | ttft 24ms | vram - GB | ctx 2,048
  M3 Ultra 512GB                     | 4-bit | 100 tok/s      | reported | ttft 79ms | vram - GB | ctx 8,192
  Radeon AI Pro R9700 32GB ×3        | 8-bit | 79 tok/s       | reported | ttft 204ms | vram - GB | ctx 686
  RX 7900 XTX 24GB                   | 4-bit | 43.8 tok/s     | reported | ttft 2,116ms | vram 24.2 GB | ctx 2,048
  RX 7900 XTX 24GB ×2                | 8-bit | 126.9 tok/s    | reported | ttft 191ms | vram - GB | ctx 32,768
  Ryzen AI Max 395 128GB             | 5-bit | 80.4 tok/s     | measured | ttft -ms | vram 23.9 GB | ctx 100,000
  Ryzen AI Max 395 64GB              | 4-bit | 31.9 tok/s     | measured | ttft 699ms | vram 30.3 GB | ctx 131,072
  Tesla V100 32GB ×2                 | 6-bit | 100.5 tok/s    | reported | ttft -ms | vram 58 GB | ctx 65,536