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=rtx-3090-24gb-x2.
Full machine surface: /llms.txt · REST API: api.bestmodel.run

  rig                               | model                                  | basis    | value          | n
  rtx-3090-24gb-x2                   | liquidai-lfm2-5-8b-a1b-gguf            | measured | 276.3 tok/s    | n=4
  rtx-3090-24gb-x2                   | nvidia-qwen3-6-35b-a3b-nvfp4           | measured | 233.4 tok/s    | n=11
  rtx-3090-24gb-x2                   | nvidia-nvidia-nemotron-3-5-lightning-30b-a3b-nvfp4 | reported | 197 tok/s      | n=2
  rtx-3090-24gb-x2                   | dbirks-qwen3-8-27b-w4a16-autoround     | measured | 182.4 tok/s    | n=3
  rtx-3090-24gb-x2                   | redhatai-diffusiongemma-26b-a4b-it-fp8-dynamic | reported | 177.6 tok/s    | n=1
  rtx-3090-24gb-x2                   | qwen-qwen3-6-35b-a3b                   | measured | 155.6 tok/s    | n=5
  rtx-3090-24gb-x2                   | google-gemma-4-26b-a4b-it              | reported | 154.1 tok/s    | n=2
  rtx-3090-24gb-x2                   | qwen-qwen3-6-35b-a3b-fp8               | reported | 140 tok/s      | n=1
  rtx-3090-24gb-x2                   | djlougen-ornstein3-6-35b-a3b-gguf      | reported | 138 tok/s      | n=1
  rtx-3090-24gb-x2                   | qwen-qwen3-6-35b-a3b                   | reported | 137.9 tok/s    | n=1
  rtx-3090-24gb-x2                   | ornith-ai-ornith-1-0-35b-gguf          | reported | 136.5 tok/s    | n=1
  rtx-3090-24gb-x2                   | google-gemma-4-31b-it                  | reported | 135.1 tok/s    | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-6-35b-a3b-gguf           | reported | 125.9 tok/s    | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-6-35b-a3b-gguf           | reported | 124 tok/s      | n=1
  rtx-3090-24gb-x2                   | turbulencedeterministe-carnice-9b-w8a16-awq | reported | 114.3 tok/s    | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-5-9b-gguf                | reported | 110.1 tok/s    | n=1
  rtx-3090-24gb-x2                   | gestaltlabs-qwen3-6-35b-a3b-nsc-ace-saber-gguf | reported | 106.2 tok/s    | n=1
  rtx-3090-24gb-x2                   | btbtyler09-qwen3-8-27b-gptq-4bit       | reported | 96.6 tok/s     | n=1
  rtx-3090-24gb-x2                   | google-gemma-4-31b-it                  | reported | 87.7 tok/s     | n=1
  rtx-3090-24gb-x2                   | lorbus-qwen3-6-27b-int4-autoround      | measured | 78.2 tok/s     | n=7
  rtx-3090-24gb-x2                   | unsloth-qwen3-8-27b-gguf               | reported | 73 tok/s       | n=1
  rtx-3090-24gb-x2                   | qwen-qwen3-8-27b                       | measured | 72.8 tok/s     | n=3
  rtx-3090-24gb-x2                   | qwen-qwen3-5-9b                        | reported | 71.7 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-5-9b-mtp-gguf            | reported | 70.6 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-6-27b-mtp-gguf           | measured | 65.2 tok/s     | n=7
  rtx-3090-24gb-x2                   | bottlecapai-thinkingcap-qwen3-6-27b    | measured | 58.6 tok/s     | n=6
  rtx-3090-24gb-x2                   | hampsonw-qwen3-6-27b-awq-bf16-int4-mtp-bf16 | reported | 57.5 tok/s     | n=2
  rtx-3090-24gb-x2                   | jackrong-qwopus3-6-27b-v2-gguf         | reported | 56.4 tok/s     | n=1
  rtx-3090-24gb-x2                   | anonimousa-qwen3-8-flash-next-reap-256-duo-gguf | reported | 53.6 tok/s     | n=2
  rtx-3090-24gb-x2                   | unsloth-qwen3-8-27b-gguf               | measured | 51.6 tok/s     | n=5
  rtx-3090-24gb-x2                   | gestaltlabs-ornstein3-6-27b-mtp-nsc-ace-saber-gguf | reported | 49.5 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-gemma-4-12b-it-gguf            | measured | 48.7 tok/s     | n=3
  rtx-3090-24gb-x2                   | gestaltlabs-ornstein3-6-27b-mtp-nsc-ace-saber-gguf | measured | 47.2 tok/s     | n=3
  rtx-3090-24gb-x2                   | qwen-qwen3-6-27b                       | reported | 46.7 tok/s     | n=1
  rtx-3090-24gb-x2                   | qwen-qwen3-6-27b-fp8                   | measured | 46.3 tok/s     | n=4
  rtx-3090-24gb-x2                   | intel-qwen3-5-27b-int4-autoround       | reported | 45.6 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-6-27b                    | reported | 45.5 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-laguna-s-2-1-gguf              | reported | 43.6 tok/s     | n=2
  rtx-3090-24gb-x2                   | cyankiwi-qwen3-6-27b-awq-bf16-int4     | reported | 43.2 tok/s     | n=2
  rtx-3090-24gb-x2                   | unsloth-qwen3-8-flash-next-gguf        | reported | 41.6 tok/s     | n=2
  rtx-3090-24gb-x2                   | unsloth-qwen3-6-27b                    | measured | 41.2 tok/s     | n=4
  rtx-3090-24gb-x2                   | qwen-qwen3-6-27b                       | reported | 40.9 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-6-35b-a3b-gguf           | reported | 40.3 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-6-27b-gguf               | measured | 37.8 tok/s     | n=3
  rtx-3090-24gb-x2                   | unsloth-muse-glimmer-30b-gguf          | reported | 31.7 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-6-27b-gguf               | reported | 23 tok/s       | n=2
  rtx-3090-24gb-x2                   | qwen-qwen3-6-27b                       | reported | 22.7 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-qwen3-6-27b-gguf               | reported | 22.2 tok/s     | n=1
  rtx-3090-24gb-x2                   | unsloth-gemma-4-31b-it-gguf            | reported | 14.4 tok/s     | n=1

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