⌕Is ___ SOTA?
benchmarks / livebench

LiveBench

score. Status active. Contamination risk low. Maintainer LiveBench / Epoch collection. Source.

ModelScoreDateKind
Gemini 2.5 Pro Exp 03.2582.352025-03-25observation
Claude 3.7 Sonnet76.12025-02-24observation
O3 Mini75.882025-01-31observation
O175.672024-12-17observation
Qwq 32B71.962025-03-05observation
DeepSeek R171.572025-01-20observation
O3 Mini70.012025-01-31observation
GPT-4.5 Preview68.952025-02-27observation
Gemini 2.0 Flash Thinking Exp 01.2166.922025-01-21observation
DeepSeek V3 032466.862025-03-24observation
Gemini 2.0 Pro Exp 02.0565.132025-02-05observation
Gemini Exp 120664.092024-12-06observation
O3 Mini62.452025-01-31observation
Qwen2 562.292025-01-28observation
Gemini 2.0 Flash 00161.472025-02-05observation
DeepSeek V360.452024-12-26observation
Gemini 2.0 Flash Exp59.262024-12-11observation
Claude 3.5 Sonnet59.032024-10-22observation
O1 Mini57.762024-09-12observation
GPT-4O52.192024-11-20observation
DeepSeek R1 Distill Llama 70B54.52025-01-20observation
Grok 2 121254.32024-12-12observation
Gemini 2.0 Flash Lite54.292025-02-05observation
Gemini 2.0 Flash Lite Preview 02.0553.242025-02-05observation
Dracarys2 72B Instruct52.642024-09-30observation
Llama 3.3 70B Instruct50.162024-12-06observation
Gemma 3 27B IT49.992025-03-12observation
Claude 3 Opus49.162024-02-29observation
Mistral Large 241148.432024-11-18observation
Qwen2 5 Coder 32B Instruct46.232024-11-21observation