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=ryzen-ai-max-395-128gb. Full machine surface: /llms.txt · REST API: api.bestmodel.run rig | model | basis | value | n ryzen-ai-max-395-128gb | jcbtc-qwen3-6-27b-rocmfp4-ciru-chaos-speed | reported | 132.3 tok/s | n=1 ryzen-ai-max-395-128gb | qwen-qwen3-6-27b | reported | 112 tok/s | n=1 ryzen-ai-max-395-128gb | google-gemma-4-e2b-it | measured | 107 tok/s | n=3 ryzen-ai-max-395-128gb | qwen-qwen3-coder-30b-a3b-instruct | measured | 80.4 tok/s | n=5 ryzen-ai-max-395-128gb | unsloth-qwen3-8-flash-next-gguf | measured | 74.9 tok/s | n=5 ryzen-ai-max-395-128gb | samuelcardillo-carnice-moe-35b-a3b | measured | 72.6 tok/s | n=4 ryzen-ai-max-395-128gb | samuelcardillo-qwopus-moe-35b-a3b | measured | 72.5 tok/s | n=6 ryzen-ai-max-395-128gb | kwaipilot-kat-coder-v2-5-dev | reported | 70.3 tok/s | n=2 ryzen-ai-max-395-128gb | unsloth-qwen3-coder-30b-a3b-instruct-gguf | measured | 70.3 tok/s | n=3 ryzen-ai-max-395-128gb | jcbtc-qwen3-8-27b-iu4-kairic-edge | measured | 66.9 tok/s | n=3 ryzen-ai-max-395-128gb | jiunsong-supergemma4-26b-uncensored-gguf-v2 | reported | 66.1 tok/s | n=1 ryzen-ai-max-395-128gb | qwen-qwen3-8-27b | measured | 65 tok/s | n=5 ryzen-ai-max-395-128gb | hauhaucs-qwen3-6-35b-a3b-uncensored-hauhaucs-aggressive | reported | 64.9 tok/s | n=2 ryzen-ai-max-395-128gb | jackrong-gemopus-4-26b-a4b-it | reported | 64.2 tok/s | n=2 ryzen-ai-max-395-128gb | internscience-agents-a1-q4-k-m-gguf | reported | 62 tok/s | n=1 ryzen-ai-max-395-128gb | qwen-qwen3-5-4b | measured | 61.3 tok/s | n=3 ryzen-ai-max-395-128gb | google-gemma-4-26b-a4b-it | measured | 60.7 tok/s | n=4 ryzen-ai-max-395-128gb | unsloth-qwen3-6-35b-a3b | measured | 57.5 tok/s | n=5 ryzen-ai-max-395-128gb | google-gemma-4-e4b-it | measured | 57.3 tok/s | n=3 ryzen-ai-max-395-128gb | qwen-qwen3-coder-next | measured | 55.8 tok/s | n=4 ryzen-ai-max-395-128gb | google-gemma-4-26b-a4b-it | reported | 55.5 tok/s | n=1 ryzen-ai-max-395-128gb | samuelcardillo-qwopus-moe-35b-a3b-gguf | reported | 54.7 tok/s | n=2 ryzen-ai-max-395-128gb | unsloth-qwen3-6-35b-a3b-mtp-gguf | measured | 54.5 tok/s | n=8 ryzen-ai-max-395-128gb | ornith-ai-ornith-1-0-35b | reported | 54.1 tok/s | n=1 ryzen-ai-max-395-128gb | mia-ailab-qwable-3-6-35b | reported | 54.1 tok/s | n=1 ryzen-ai-max-395-128gb | qwen-qwen3-8-flash-next | reported | 54 tok/s | n=1 ryzen-ai-max-395-128gb | samuelcardillo-qwopus-moe-35b-a3b | measured | 53.1 tok/s | n=4 ryzen-ai-max-395-128gb | unsloth-qwen3-6-35b-a3b | measured | 52.8 tok/s | n=3 ryzen-ai-max-395-128gb | unsloth-glm-4-7-flash-gguf | reported | 52.8 tok/s | n=2 ryzen-ai-max-395-128gb | jcbtc-qwen3-6-27b-rocmfp4-ciru-chaos-speed | reported | 52.2 tok/s | n=1 ryzen-ai-max-395-128gb | qwen-qwen3-coder-next | measured | 51.2 tok/s | n=5 ryzen-ai-max-395-128gb | google-gemma-4-26b-a4b-it | reported | 46.1 tok/s | n=2 ryzen-ai-max-395-128gb | jackrong-gemopus-4-26b-a4b-it | reported | 45.8 tok/s | n=2 ryzen-ai-max-395-128gb | ornith-ai-ornith-1-0-35b-gguf | reported | 43 tok/s | n=2 ryzen-ai-max-395-128gb | lmcoleman-qwen3-8-27b-rocmfpx-gguf | reported | 41 tok/s | n=1 ryzen-ai-max-395-128gb | poolside-laguna-s-2-1 | measured | 40.7 tok/s | n=13 ryzen-ai-max-395-128gb | unsloth-qwen3-8b-gguf | reported | 39.3 tok/s | n=1 ryzen-ai-max-395-128gb | inclusionai-ling-3-0-flash | reported | 39.2 tok/s | n=1 ryzen-ai-max-395-128gb | openai-gpt-oss-120b | reported | 38.8 tok/s | n=1 ryzen-ai-max-395-128gb | nousresearch-hermes-3-llama-3-1-8b | measured | 38.5 tok/s | n=3 ryzen-ai-max-395-128gb | qwen-qwen3-5-9b | measured | 37.2 tok/s | n=3 ryzen-ai-max-395-128gb | empero-ai-qwythos-9b-claude-mythos-5-1m-gguf | reported | 35.3 tok/s | n=1 ryzen-ai-max-395-128gb | unsloth-qwen-agentworld-35b-a3b-gguf | reported | 34.9 tok/s | n=2 ryzen-ai-max-395-128gb | prism-ml-bonsai-27b-gguf | measured | 33.6 tok/s | n=4 ryzen-ai-max-395-128gb | ornith-ai-ornith-1-0-9b-gguf | reported | 31.8 tok/s | n=1 ryzen-ai-max-395-128gb | lucebox-deepseek-v4-flash-0731-rocmfp3 | reported | 30.1 tok/s | n=2 ryzen-ai-max-395-128gb | deepseek-ai-deepseek-v4-flash-0731 | reported | 28.3 tok/s | n=2 ryzen-ai-max-395-128gb | xiaomimimo-mimo-v2-5 | reported | 28.1 tok/s | n=1 ryzen-ai-max-395-128gb | unsloth-qwen3-8-27b-gguf | measured | 26.7 tok/s | n=5 ryzen-ai-max-395-128gb | hauhaucs-gemma4-12b-qat-uncensored-hauhaucs-balanced | reported | 24.8 tok/s | n=1 ryzen-ai-max-395-128gb | yuxinlu1-gemma-4-12b-coder-fable5-composer2-5-v1-gguf | reported | 24.8 tok/s | n=1 ryzen-ai-max-395-128gb | yuxinlu1-gemma-4-12b-agentic-fable5-composer2-5-v2-3-5x-tau2-gguf | reported | 24.7 tok/s | n=1 ryzen-ai-max-395-128gb | kingjones777-qwen3-8-27b-rocmfp4-strix-mtp-gguf | reported | 23.9 tok/s | n=1 ryzen-ai-max-395-128gb | prism-ml-ternary-bonsai-27b-gguf | measured | 23.2 tok/s | n=4 ryzen-ai-max-395-128gb | empero-ai-qwythos-9b-claude-mythos-5-1m-gguf | reported | 22.3 tok/s | n=2 ryzen-ai-max-395-128gb | stepfun-ai-step-3-7-flash-gguf | reported | 22 tok/s | n=1 ryzen-ai-max-395-128gb | deepseek-ai-deepseek-v4-flash-0731 | reported | 21.8 tok/s | n=1 ryzen-ai-max-395-128gb | qwen-qwen3-5-122b-a10b | reported | 18 tok/s | n=2 ryzen-ai-max-395-128gb | unsloth-qwen3-6-27b-mtp-gguf | measured | 17.1 tok/s | n=3 ryzen-ai-max-395-128gb | qwen-qwen3-6-27b | reported | 16.7 tok/s | n=1 Capture or correct a number: /console (signed, Ed25519).