Local LLM benchmarks

Every number here was pasted by the person who ran it, and every page keeps the original output it came from. Ran something yourself? Paste a benchmark — no account needed.

• Qwen3.8 27B18.75 tok/s276.46 promptQ8_0 · 2× AMD Instinct MI60 / MI50 · KV q8_0 · llama.cppSep 2
• Qwen3.8 27B18.67 tok/s274.95 promptQ8_0 · 2× AMD Instinct MI60 / MI50 · KV q8_0 · llama.cppSep 2
• Deepseek V4 Flash 073160 tok/sprompt not measuredNVFP4 · 2× DGX Spark 128GB · vLLMAug 26
• Qwen3.8 Flash Next31.67 tok/s366.79 promptIQ1_S (UD) · 2× AMD Mi50 32GB · KV q8_0 · llama.cppAug 30
• Qwen3.8 27B30.33 tok/s512.67 promptQ8_0 · 2× AMD Mi50 32GB · KV q4_0 · llama.cppAug 29