bestmodel.run / model — Qwen3.5-27B (agent view)

category chat · Qwen/Qwen3.5-27B · 18 runs in the model record
answer: 25.1 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                   | 4-bit | 18 tok/s       | reported | ttft 486ms | vram 20 GB | ctx 8,192
  Arc Pro B70 32GB                   | 5-bit | 25.1 tok/s     | reported | ttft -ms | vram - GB | ctx 204,800
  GB10 Grace Blackwell 128GB         | 4-bit | 11.5 tok/s     | measured | ttft 658ms | vram 113.5 GB | ctx 20,000
  GTX 1080 Ti 11GB                   | 4-bit | 2.5 tok/s      | reported | ttft -ms | vram 10.3 GB | ctx 8,192
  Radeon 8060S Graphics 103GB        | 4-bit | 104.5 tok/s    | reported | ttft 338ms | vram 18 GB | ctx 4,096
  Radeon 8060S Graphics (Strix Halo APU, gfx1151) 96GB | 4-bit | 14.8 tok/s     | reported | ttft 480ms | vram 16 GB | ctx 65,536
  Radeon AI Pro R9700 32GB           | 4-bit | 196.2 tok/s    | reported | ttft 133ms | vram 17.9 GB | ctx 4,096
  Radeon AI Pro R9700 34GB           | 4-bit | 286.6 tok/s    | reported | ttft 111ms | vram 17.5 GB | ctx 4,096
  RX 7900 XTX 24GB                   | 4-bit | 191.5 tok/s    | measured | ttft 136ms | vram 17 GB | ctx 4,096