bestmodel.run / model — Qwen3-Coder-30B-A3B-Instruct-4bit (agent view)

category code · mlx-community/Qwen3-Coder-30B-A3B-Instruct-4bit · 2 runs in the model record
answer: 120.9 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
  M5 Max 128GB                       | 4-bit | 140.7 tok/s    | reported | ttft -ms | vram 17.5 GB | ctx 2,048
  M5 Pro 64GB                        | 4-bit | 101.1 tok/s    | reported | ttft 222ms | vram 17.5 GB | ctx 2,048