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=arc-pro-b70-32gb. Full machine surface: /llms.txt · REST API: api.bestmodel.run rig | model | basis | value | n arc-pro-b70-32gb | palmfuture-qwen3-6-35b-a3b-gptq-int4 | reported | 248.3 tok/s | n=1 arc-pro-b70-32gb | redhatai-qwen3-5-4b-quantized-w4a16 | reported | 184.5 tok/s | n=2 arc-pro-b70-32gb | liquidai-lfm2-5-2-6b-gguf | reported | 132.4 tok/s | n=1 arc-pro-b70-32gb | ornith-ai-ornith-1-5-35b-a3b-gguf | reported | 131.5 tok/s | n=1 arc-pro-b70-32gb | redhatai-qwen3-5-9b-quantized-w4a16 | reported | 118.7 tok/s | n=2 arc-pro-b70-32gb | unsloth-qwen3-coder-30b-a3b-instruct-gguf | reported | 108.1 tok/s | n=1 arc-pro-b70-32gb | unsloth-qwen3-30b-a3b-instruct-2507-gguf | reported | 107.5 tok/s | n=1 arc-pro-b70-32gb | nex-agi-nex-n2-mini | measured | 104.5 tok/s | n=7 arc-pro-b70-32gb | qwen-qwen3-4b | reported | 100.5 tok/s | n=2 arc-pro-b70-32gb | qwen-qwen3-30b-a3b-instruct-2507 | reported | 100.5 tok/s | n=1 arc-pro-b70-32gb | redhatai-qwen3-5-9b-fp8-dynamic | reported | 98.1 tok/s | n=1 arc-pro-b70-32gb | johannrplaster-qwen3-8-27b-uncensored-int4-autoround | reported | 97.8 tok/s | n=1 arc-pro-b70-32gb | unsloth-gemma-4-26b-a4b-it-gguf | measured | 97.6 tok/s | n=42 arc-pro-b70-32gb | bartowski-microsoft-phi-4-mini-instruct-gguf | reported | 96.6 tok/s | n=1 arc-pro-b70-32gb | sergiiob-ornith-1-5-35b-a3b-gptq-int4-sym-g128-mtp-bf16-mixedcal-v2 | measured | 94.1 tok/s | n=3 arc-pro-b70-32gb | nvidia-nvidia-nemotron-3-5-lightning-30b-a3b-bf16 | measured | 93 tok/s | n=9 arc-pro-b70-32gb | poolside-laguna-xs-2-1 | reported | 87.8 tok/s | n=1 arc-pro-b70-32gb | bartowski-nvidia-nemotron-3-5-lightning-30b-a3b-gguf | reported | 87.4 tok/s | n=1 arc-pro-b70-32gb | sergiiob-qwen3-8-27b-gptq-int4-sym-g128-mtp-bf16 | measured | 80.6 tok/s | n=6 arc-pro-b70-32gb | ornith-ai-ornith-1-0-35b | reported | 74.3 tok/s | n=2 arc-pro-b70-32gb | internscience-agents-a1 | reported | 73.2 tok/s | n=2 arc-pro-b70-32gb | bartowski-microsoft-phi-4-mini-instruct-gguf | reported | 72.3 tok/s | n=1 arc-pro-b70-32gb | unsloth-nvidia-nemotron-3-5-lightning-30b-a3b-gguf | reported | 72.2 tok/s | n=1 arc-pro-b70-32gb | qwen-qwen3-6-35b-a3b | measured | 70 tok/s | n=8 arc-pro-b70-32gb | ornith-ai-ornith-1-0-35b | measured | 69.3 tok/s | n=3 arc-pro-b70-32gb | qwen-qwen3-6-35b-a3b | measured | 67 tok/s | n=3 arc-pro-b70-32gb | webhie-qwen3-6-27b-int4-autoround | measured | 66.4 tok/s | n=4 arc-pro-b70-32gb | frosty40-qwen-agentworld-35b-a3b-b70-turbo-gguf | measured | 66 tok/s | n=4 arc-pro-b70-32gb | frosty40-nex-n2-mini-b70-turbo-gguf | measured | 65.8 tok/s | n=4 arc-pro-b70-32gb | frosty40-siq-1-35b-b70-turbo-gguf | measured | 65.6 tok/s | n=4 arc-pro-b70-32gb | qwen-qwen3-5-35b-a3b | reported | 62.8 tok/s | n=1 arc-pro-b70-32gb | qwen-qwen3-5-35b-a3b | reported | 61.5 tok/s | n=1 arc-pro-b70-32gb | frozenlock-qwen3-8-27b-int4-autoround | reported | 60.2 tok/s | n=1 arc-pro-b70-32gb | hauhaucs-gemma4-26b-a4b-qat-uncensored-hauhaucs-balanced-mtp | reported | 58.4 tok/s | n=2 arc-pro-b70-32gb | intel-qwen3-6-27b-int4-autoround | reported | 58.1 tok/s | n=2 arc-pro-b70-32gb | bartowski-deepseek-coder-v2-lite-instruct-gguf | reported | 57.1 tok/s | n=1 arc-pro-b70-32gb | qwen-qwen3-8-27b-fp8 | measured | 54.2 tok/s | n=4 arc-pro-b70-32gb | bartowski-nvidia-nemotron-cascade-2-30b-a3b-gguf | reported | 50.9 tok/s | n=1 arc-pro-b70-32gb | ornith-ai-ornith-1-0-9b | reported | 50.3 tok/s | n=2 arc-pro-b70-32gb | ornith-ai-ornith-1-5-9b-gguf | reported | 49.6 tok/s | n=1 arc-pro-b70-32gb | qwen-qwen3-8-27b | measured | 44 tok/s | n=9 arc-pro-b70-32gb | qwen-qwen3-6-27b | measured | 43.5 tok/s | n=14 arc-pro-b70-32gb | google-gemma-4-26b-a4b-it | reported | 43.3 tok/s | n=1 arc-pro-b70-32gb | lorbus-qwen3-6-27b-int4-autoround | measured | 41.3 tok/s | n=3 arc-pro-b70-32gb | unsloth-glm-4-7-flash-gguf | reported | 40.8 tok/s | n=1 arc-pro-b70-32gb | lm-kit-qwen-3-14b-instruct-gguf | reported | 38.3 tok/s | n=1 arc-pro-b70-32gb | unsloth-qwen3-6-27b-mtp-gguf | reported | 36.1 tok/s | n=1 arc-pro-b70-32gb | unsloth-qwen3-8-27b-gguf | reported | 31.2 tok/s | n=1 arc-pro-b70-32gb | unsloth-qwen3-6-27b-mtp-gguf | reported | 30.7 tok/s | n=1 arc-pro-b70-32gb | meta-models-muse-glimmer-30b-gguf | measured | 28.8 tok/s | n=4 arc-pro-b70-32gb | bottlecapai-thinkingcap-qwen3-6-27b | reported | 27.9 tok/s | n=1 arc-pro-b70-32gb | ggml-org-qwen3-8-27b-gguf | reported | 27.8 tok/s | n=1 arc-pro-b70-32gb | aaryank-qwen3-6-27b-gguf | measured | 27.7 tok/s | n=5 arc-pro-b70-32gb | bottlecapai-thinkingcap-qwen3-6-27b | reported | 27.5 tok/s | n=1 arc-pro-b70-32gb | unsloth-mistral-small-3-2-24b-instruct-2506-gguf | reported | 27.3 tok/s | n=1 arc-pro-b70-32gb | meta-models-muse-glimmer-30b | measured | 26.8 tok/s | n=3 arc-pro-b70-32gb | qwen-qwen3-6-27b | measured | 26.6 tok/s | n=3 arc-pro-b70-32gb | bottlecapai-thinkingcap-qwen3-6-27b | reported | 26.3 tok/s | n=1 arc-pro-b70-32gb | unsloth-qwen3-6-27b-gguf | measured | 26 tok/s | n=5 arc-pro-b70-32gb | qwen-qwen3-5-27b | reported | 25.1 tok/s | n=1 Capture or correct a number: /console (signed, Ed25519).