bestmodel.run / model — KAT-Coder-V2.5-Dev (agent view)

category code · Kwaipilot/KAT-Coder-V2.5-Dev · 4 runs in the model record
answer: 75 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
  Radeon AI Pro R9700 32GB ×3        | 8-bit | 75.8 tok/s     | reported | ttft 344ms | vram 42.8 GB | ctx 787
  Ryzen AI Max 395 128GB             | 4-bit | 70.3 tok/s     | reported | ttft 1,359ms | vram - GB | ctx 787