bestmodel.run / model — Qwen2.5-Coder-14B-Instruct (agent view)

category code · Qwen/Qwen2.5-Coder-14B-Instruct · 13 runs in the model record
answer: 35 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
  RTX 2080 8GB                       | 4-bit | 11 tok/s       | reported | ttft -ms | vram 9 GB | ctx 2,048
  RTX 3060 12GB                      | 4-bit | 35 tok/s       | reported | ttft -ms | vram 9 GB | ctx 2,048
  RTX 3080 12GB                      | 4-bit | 72 tok/s       | measured | ttft -ms | vram 9 GB | ctx 2,048
  RTX 5000 16GB                      | 5-bit | 24.3 tok/s     | reported | ttft -ms | vram 14.6 GB | ctx 32,768
  RTX PRO 6000 Blackwell 96GB        | 16-bit | 50.3 tok/s     | reported | ttft -ms | vram - GB | ctx 32,768
  RTX3060+RTX2080-pooled-20GB 20GB   | 4-bit | 35 tok/s       | measured | ttft -ms | vram 9 GB | ctx 2,048