bestmodel.run / pool — agent view

Honesty ladder: measured > reported > extrapolated > formula > no data yet.
Top 60 cells, default sort: fastest decode. Columns: rig | model | basis | value | n.
Filter: rig=m5-max-128gb.
Full machine surface: /llms.txt · REST API: api.bestmodel.run

  rig                               | model                                  | basis    | value          | n
  m5-max-128gb                       | mlx-community-gemma-3-text-4b-it-4bit  | reported | 175.1 tok/s    | n=1
  m5-max-128gb                       | mlx-community-qwen3-6-35b-a3b-mxfp4    | reported | 143.3 tok/s    | n=1
  m5-max-128gb                       | mlx-community-qwen3-coder-30b-a3b-instruct-4bit | reported | 140.7 tok/s    | n=1
  m5-max-128gb                       | mlx-community-qwen3-5-35b-a3b-4bit     | reported | 140.1 tok/s    | n=1
  m5-max-128gb                       | qwen-qwen3-vl-30b-a3b-instruct-gguf    | reported | 138.6 tok/s    | n=1
  m5-max-128gb                       | openai-gpt-oss-20b                     | reported | 137 tok/s      | n=1
  m5-max-128gb                       | qwen-qwen3-5-35b-a3b                   | reported | 122 tok/s      | n=1
  m5-max-128gb                       | lmstudio-community-gemma-4-26b-a4b-it-mlx-4bit | reported | 118.8 tok/s    | n=1
  m5-max-128gb                       | lmstudio-community-nemotron-3-nano-omni-30b-a3b-reasoning-gguf | reported | 113.7 tok/s    | n=1
  m5-max-128gb                       | qwen-qwen3-6-35b-a3b                   | measured | 105.2 tok/s    | n=3
  m5-max-128gb                       | hesamation-qwen3-6-35b-a3b-claude-4-6-opus-reasoning-distilled | reported | 104.1 tok/s    | n=1
  m5-max-128gb                       | hauhaucs-qwen3-6-35b-a3b-uncensored-hauhaucs-aggressive | reported | 99.7 tok/s     | n=1
  m5-max-128gb                       | unsloth-gemma-4-26b-a4b-it-gguf        | reported | 99.2 tok/s     | n=1
  m5-max-128gb                       | unsloth-qwen3-6-35b-a3b-mtp-gguf       | measured | 97 tok/s       | n=3
  m5-max-128gb                       | lmstudio-community-gemma-4-e4b-it-mlx-4bit | reported | 91.4 tok/s     | n=1
  m5-max-128gb                       | lmstudio-community-qwen3-coder-next-mlx-4bit | reported | 85.2 tok/s     | n=1
  m5-max-128gb                       | pipenetwork-laguna-s-2-1-mlx-4bit      | measured | 63.8 tok/s     | n=5
  m5-max-128gb                       | unsloth-laguna-s-2-1-gguf              | reported | 53 tok/s       | n=1
  m5-max-128gb                       | qwen-qwen3-5-122b-a10b                 | reported | 50.1 tok/s     | n=1
  m5-max-128gb                       | unsloth-qwen3-8-flash-next-gguf        | reported | 41.5 tok/s     | n=1
  m5-max-128gb                       | mcsplain-qwen3-8-27b-oq8-mtp           | reported | 39.6 tok/s     | n=1
  m5-max-128gb                       | unsloth-qwen3-8-flash-next-gguf        | reported | 39.4 tok/s     | n=1
  m5-max-128gb                       | jackrong-qwopus3-8-27b-flash-gguf      | reported | 26.5 tok/s     | n=2
  m5-max-128gb                       | hotdogs-qwen27b-abliterated-fable-mtp  | reported | 26.2 tok/s     | n=1
  m5-max-128gb                       | unsloth-qwen3-8-27b-gguf               | reported | 24.3 tok/s     | n=1
  m5-max-128gb                       | unsloth-gemma-4-31b-it-gguf            | reported | 24.1 tok/s     | n=1
  m5-max-128gb                       | qwen-qwen3-6-27b                       | reported | 19 tok/s       | n=1
  m5-max-128gb                       | deepseek-ai-deepseek-v4-flash          | measured | 19 tok/s       | n=5
  m5-max-128gb                       | unsloth-gemma-4-31b-it-gguf            | reported | 18.1 tok/s     | n=1
  m5-max-128gb                       | lmstudio-community-gemma-4-31b-it-mlx-4bit | reported | 17.8 tok/s     | n=1
  m5-max-128gb                       | blackfrost-ai-qwen3-8-27b-abliterated-gguf | reported | 17.3 tok/s     | n=1
  m5-max-128gb                       | unsloth-gemma-4-31b-it-gguf            | reported | 15.3 tok/s     | n=1
  m5-max-128gb                       | huihui-ai-huihui-deepseek-v4-flash-abliterated-ds4-gguf | reported | 12.1 tok/s     | n=1
  m5-max-128gb                       | unsloth-gemma-4-31b-it-gguf            | reported | 8.9 tok/s      | n=1

Capture or correct a number: /console (signed, Ed25519).