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-3060-12gb. Full machine surface: /llms.txt · REST API: api.bestmodel.run rig | model | basis | value | n rtx-3060-12gb | qwen-qwen3-0-6b | reported | 310.4 tok/s | n=1 rtx-3060-12gb | liquidai-lfm2-5-1-2b-instruct | reported | 231.2 tok/s | n=1 rtx-3060-12gb | liquidai-lfm2-8b-a1b | measured | 225.7 tok/s | n=21 rtx-3060-12gb | xhtoken-spark-x2-5-1-7b | reported | 206.7 tok/s | n=2 rtx-3060-12gb | liquidai-lfm2-5-8b-a1b | measured | 194.4 tok/s | n=20 rtx-3060-12gb | xhtoken-spark-x2-5-1-7b | reported | 166.4 tok/s | n=1 rtx-3060-12gb | qwen-qwen3-5-2b | reported | 166.2 tok/s | n=2 rtx-3060-12gb | xhtoken-spark-x2-5-1-7b | reported | 149.6 tok/s | n=1 rtx-3060-12gb | deepseek-ai-deepseek-coder-v2-lite-instruct | reported | 149 tok/s | n=1 rtx-3060-12gb | qwen-qwen3-1-7b | reported | 147.7 tok/s | n=1 rtx-3060-12gb | hauhaucs-gemma-4-e2b-uncensored-hauhaucs-aggressive | reported | 133.6 tok/s | n=1 rtx-3060-12gb | google-gemma-4-e2b-it | reported | 133.4 tok/s | n=1 rtx-3060-12gb | liquidai-lfm2-5-8b-a1b | reported | 128.5 tok/s | n=1 rtx-3060-12gb | qwen-qwen2-5-coder-7b-instruct | reported | 119 tok/s | n=1 rtx-3060-12gb | qwen-qwen3-0-6b-gguf | measured | 117.5 tok/s | n=27 rtx-3060-12gb | unsloth-gemma-4-e2b-it-gguf | reported | 113.4 tok/s | n=1 rtx-3060-12gb | qwen-qwen3-30b-a3b | measured | 107 tok/s | n=3 rtx-3060-12gb | inclusionai-ling-3-0-tiny | reported | 104.8 tok/s | n=2 rtx-3060-12gb | google-gemma-3-4b-it | reported | 97 tok/s | n=1 rtx-3060-12gb | qwen-qwen3-1-7b-gguf | measured | 96.2 tok/s | n=9 rtx-3060-12gb | meta-llama-llama-3-2-3b-instruct | measured | 95.8 tok/s | n=39 rtx-3060-12gb | nvidia-nvidia-nemotron-3-nano-4b-bf16 | measured | 93.2 tok/s | n=27 rtx-3060-12gb | xhtoken-spark-x2-5-4b | reported | 90.7 tok/s | n=2 rtx-3060-12gb | mradermacher-huihui-qwythos-9b-claude-mythos-5-1m-abliterated-i1-gguf | reported | 90.2 tok/s | n=1 rtx-3060-12gb | lmstudio-community-nvidia-nemotron-3-nano-4b-gguf | reported | 89.7 tok/s | n=1 rtx-3060-12gb | xhtoken-spark-x2-5-4b | reported | 89.2 tok/s | n=1 rtx-3060-12gb | google-gemma-4-e4b-it-qat-q4-0-gguf | reported | 88.1 tok/s | n=1 rtx-3060-12gb | hauhaucs-qwen3-5-4b-uncensored-hauhaucs-aggressive | reported | 82.9 tok/s | n=1 rtx-3060-12gb | unsloth-gemma-4-12b-it-qat-gguf | reported | 79 tok/s | n=1 rtx-3060-12gb | unsloth-qwen3-5-4b-gguf | reported | 78.6 tok/s | n=1 rtx-3060-12gb | mistralai-ministral-3-3b-instruct-2512 | reported | 78 tok/s | n=1 rtx-3060-12gb | xhtoken-spark-x2-5-4b | reported | 77.9 tok/s | n=1 rtx-3060-12gb | hauhaucs-gemma-4-e4b-uncensored-hauhaucs-aggressive | reported | 77.5 tok/s | n=1 rtx-3060-12gb | google-gemma-4-12b-it | measured | 76.7 tok/s | n=8 rtx-3060-12gb | internscience-agents-a1-4b-q4-k-m-gguf | measured | 76.6 tok/s | n=9 rtx-3060-12gb | qwen-qwen3-5-4b | measured | 76.5 tok/s | n=9 rtx-3060-12gb | empero-ai-qwen3-8-9b-distill-gguf | reported | 76.4 tok/s | n=1 rtx-3060-12gb | xhtoken-spark-x2-5-4b | reported | 76.1 tok/s | n=1 rtx-3060-12gb | meta-llama-llama-3-1-8b-instruct | reported | 76 tok/s | n=1 rtx-3060-12gb | openai-gpt-oss-20b | reported | 74 tok/s | n=1 rtx-3060-12gb | google-gemma-4-e4b-it | reported | 73.3 tok/s | n=2 rtx-3060-12gb | mistralai-ministral-3-3b-instruct-2512 | measured | 73.2 tok/s | n=27 rtx-3060-12gb | google-gemma-4-e4b | reported | 70.5 tok/s | n=2 rtx-3060-12gb | unsloth-gemma-4-e4b-it-gguf | measured | 69.9 tok/s | n=10 rtx-3060-12gb | unsloth-phi-4-mini-reasoning-gguf | measured | 65.2 tok/s | n=27 rtx-3060-12gb | ornith-ai-ornith-1-5-9b-gguf | reported | 65 tok/s | n=1 rtx-3060-12gb | unsloth-qwen3-6-35b-a3b-gguf | measured | 63.4 tok/s | n=4 rtx-3060-12gb | deepseek-ai-deepseek-r1-distill-qwen-7b | measured | 62.1 tok/s | n=11 rtx-3060-12gb | lmstudio-community-openreasoning-nemotron-7b-gguf | measured | 57.3 tok/s | n=27 rtx-3060-12gb | nvidia-acereason-nemotron-1-1-7b | measured | 57.3 tok/s | n=9 rtx-3060-12gb | ibm-granite-granite-4-2-8b-gguf | reported | 55.3 tok/s | n=2 rtx-3060-12gb | jackrong-qwopus3-5-9b-coder-mtp-gguf | reported | 53.8 tok/s | n=1 rtx-3060-12gb | ibm-granite-granite-4-2-8b-gguf | reported | 52.5 tok/s | n=1 rtx-3060-12gb | ornith-ai-ornith-1-0-9b-gguf | reported | 52.5 tok/s | n=1 rtx-3060-12gb | jackrong-qwen3-5-9b-claude-4-6-opus-reasoning-distilled-v2-gguf | reported | 52.4 tok/s | n=1 rtx-3060-12gb | unsloth-qwen3-5-9b-gguf | reported | 52.2 tok/s | n=2 rtx-3060-12gb | kai-os-carnice-9b-gguf | measured | 52 tok/s | n=10 rtx-3060-12gb | qwen-qwen3-5-9b | measured | 51.4 tok/s | n=9 rtx-3060-12gb | empero-ai-qwythos-9b-v2-gguf | reported | 51.1 tok/s | n=1 rtx-3060-12gb | weiboai-vibethinker-3b | measured | 45.9 tok/s | n=18 Capture or correct a number: /console (signed, Ed25519).