bestmodel.run / model — Qwen3.6-35B-A3B-FP8 (agent view)

category chat · Qwen/Qwen3.6-35B-A3B-FP8 · 86 runs in the model record
answer: 161.7 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
  B200 192GB                         | 8-bit | 220.4 tok/s    | reported | ttft 28ms | vram 161.3 GB | ctx 32,768
  CMP 170HX 64GB                     | 8-bit | 155.7 tok/s    | measured | ttft 1,868ms | vram 55.8 GB | ctx 131,335
  GB10 Grace Blackwell 121GB         | 8-bit | 54.1 tok/s     | reported | ttft -ms | vram - GB | ctx 128,000
  H100 80GB                          | 8-bit | 337.1 tok/s    | reported | ttft 58ms | vram 71.6 GB | ctx 32,768
  H200 SXM 141GB                     | 8-bit | 261.5 tok/s    | reported | ttft 48ms | vram 126.7 GB | ctx 32,768
  RTX 3090 24GB ×2                   | 8-bit | 140 tok/s      | reported | ttft 149ms | vram 22.2 GB | ctx 32,768
  RTX PRO 6000 Blackwell 96GB        | 8-bit | 253.7 tok/s    | reported | ttft 229ms | vram 38.4 GB | ctx 4,096