Skip to content
AI Atlas
BenchmarkActivecategory · general

LiveBench

livebench.ai

contamination-limited, monthly refreshed questions across 6 categories

quality57

Updated 8 h ago · first seen 11 Sept 2026

bench_01M293SPERF0D8QG7TSCGE8GGF

Metric
average score · %
Direction
Higher is better
Results
456
Leader
Claude Fable 5.1 Max Effort 97.01%

Score history · claude-sonnet-4-6-thinking-auto-medium-effort 8 rows

Not enough history to chart — 8 observations, all dated 25 Jun 2026. Rows under different configurations count separately; the list below shows each one.

  • 63.22%release=2026-06-25 · subtasks=["paraphrase","simplify","story_generation","summarize"] · livebench_model_id=claude-sonnet-4-6-thinking-auto-medium-effort25 Jun 2026
  • 76.1%release=2026-06-25 · subtasks=["connections","plot_unscrambling","typos"] · livebench_model_id=claude-sonnet-4-6-thinking-auto-medium-effort25 Jun 2026
  • 77.95%release=2026-06-25 · subtasks=["consecutive_events","tablejoin","tablereformat"] · livebench_model_id=claude-sonnet-4-6-thinking-auto-medium-effort25 Jun 2026
  • 86.99%release=2026-06-25 · subtasks=["AMPS_Hard","integrals_with_game","math_comp","olympiad"] · livebench_model_id=claude-sonnet-4-6-thinking-auto-medium-effort25 Jun 2026
  • 42.63%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=claude-sonnet-4-6-thinking-auto-medium-effort25 Jun 2026
  • 79.27%release=2026-06-25 · subtasks=["code_generation","code_completion"] · livebench_model_id=claude-sonnet-4-6-thinking-auto-medium-effort25 Jun 2026
  • 84.77%release=2026-06-25 · subtasks=["theory_of_mind","zebra_puzzle","spatial","logic_with_navigation"] · livebench_model_id=claude-sonnet-4-6-thinking-auto-medium-effort25 Jun 2026
  • 72.99%release=2026-06-25 · aggregation=mean of category averages; category = mean of its subtasks · livebench_model_id=claude-sonnet-4-6-thinking-auto-medium-effort25 Jun 2026

Back to the full leaderboard

Leaderboard 456 current results

Select models with +, then open Compare.

Leaderboard
#ModelScoreConfigEvaluatedSourceActions
401#401ox-alpha-max60.25%release=2026-06-25 · subtasks=["paraphrase","simplify","story_generation","summarize"] · livebench_model_id=ox-alpha-max25 Jun 2026livebench.aiT2 History
402#402Gemini 3.5 Flash-Lite HighGoogle60.19%release=2026-06-25 · subtasks=["theory_of_mind","zebra_puzzle","spatial","logic_with_navigation"] · livebench_model_id=gemini-3.5-flash-lite-high25 Jun 2026livebench.aiT2 History
403#403gpt-5.6-lunaOpenAI60.12%release=2026-06-25 · subtasks=["paraphrase","simplify","story_generation","summarize"] · livebench_model_id=gpt-5.6-luna-max25 Jun 2026livebench.aiT2 History
404#404gpt-5.4-miniOpenAI59.8%release=2026-06-25 · subtasks=["paraphrase","simplify","story_generation","summarize"] · livebench_model_id=gpt-5.4-mini-xhigh25 Jun 2026livebench.aiT2 History
405#405Claude Sonnet 5 xHigh EffortAnthropic59.39%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=claude-sonnet-5-xhigh-effort25 Jun 2026livebench.aiT2 History
406#406Muse Spark 1.1Meta AI58.54%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=muse-spark-1.1-xhigh25 Jun 2026livebench.aiT2 History
407#407Qwen3.6 PlusQwen58.34%release=2026-06-25 · subtasks=["paraphrase","simplify","story_generation","summarize"] · livebench_model_id=qwen3.6-plus25 Jun 2026livebench.aiT2 History
408#408Gemini 3.7 FlashGoogle58.28%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gemini-3.7-flash-high25 Jun 2026livebench.aiT2 History
409#409Muse Spark 1.2Meta AI57.58%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=muse-spark-1.2-xhigh25 Jun 2026livebench.aiT2 History
410#410MiniMax M3MiniMax57.51%release=2026-06-25 · subtasks=["paraphrase","simplify","story_generation","summarize"] · livebench_model_id=minimax-m325 Jun 2026livebench.aiT2 History
411#411gpt-6-astraOpenAI57.32%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-6-astra-max25 Jun 2026livebench.aiT2 History
412#412Grok 4.6xAI57.02%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=grok-4.625 Jun 2026livebench.aiT2 History
413#413GLM 5.3 FlashZ.ai (Zhipu AI)56.77%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=glm-5.3-flash25 Jun 2026livebench.aiT2 History
414#414Grok 4.5xAI56.47%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=grok-4.525 Jun 2026livebench.aiT2 History
415#415Kimi K2.7 CodeMoonshot AI56.29%release=2026-06-25 · subtasks=["paraphrase","simplify","story_generation","summarize"] · livebench_model_id=kimi-k2.7-code25 Jun 2026livebench.aiT2 History
416#416gpt-5-6-solOpenAI56.21%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-5.6-sol-max25 Jun 2026livebench.aiT2 History
417#417Grok 4.3xAI55.77%release=2026-06-25 · subtasks=["consecutive_events","tablejoin","tablereformat"] · livebench_model_id=grok-4.325 Jun 2026livebench.aiT2 History
418#418gpt-5.6-terraOpenAI54.95%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-5.6-terra-max25 Jun 2026livebench.aiT2 History
419#419DeepSeek-V4-Pro-0813DeepSeek54.95%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=deepseek-v4-pro-081325 Jun 2026livebench.aiT2 History
420#420Nemotron 3 UltraNVIDIA54.46%release=2026-06-25 · subtasks=["consecutive_events","tablejoin","tablereformat"] · livebench_model_id=nemotron-3-ultra-550b-a55b25 Jun 2026livebench.aiT2 History
421#421Gemini 3.8 FlashGoogle54.24%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gemini-3.8-flash-high25 Jun 2026livebench.aiT2 History
422#422Gemini 3.8 FlashGoogle54.01%release=2026-06-25 · subtasks=["consecutive_events","tablejoin","tablereformat"] · livebench_model_id=gemini-3.8-flash-high25 Jun 2026livebench.aiT2 History
423#423gpt-5.5OpenAI53.99%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-5.5-xhigh25 Jun 2026livebench.aiT2 History
424#424gpt-5.4OpenAI53.84%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-5.4-xhigh25 Jun 2026livebench.aiT2 History
425#425Gemini 3.5 Flash-Lite HighGoogle53.25%release=2026-06-25 · subtasks=["consecutive_events","tablejoin","tablereformat"] · livebench_model_id=gemini-3.5-flash-lite-high25 Jun 2026livebench.aiT2 History
426#426Qwen3.6 27BQwen53.23%release=2026-06-25 · subtasks=["paraphrase","simplify","story_generation","summarize"] · livebench_model_id=qwen3.6-27b25 Jun 2026livebench.aiT2 History
427#427GLM 5.3 FlashZ.ai (Zhipu AI)52.82%release=2026-06-25 · subtasks=["paraphrase","simplify","story_generation","summarize"] · livebench_model_id=glm-5.3-flash25 Jun 2026livebench.aiT2 History
428#428ox-alpha-max52.63%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=ox-alpha-max25 Jun 2026livebench.aiT2 History
429#429Z.ai GLM 5.2Z.ai (Zhipu AI)51.77%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=glm-5.225 Jun 2026livebench.aiT2 History
430#430claude-opus-4-7-xhigh-effort50.66%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=claude-opus-4-7-xhigh-effort25 Jun 2026livebench.aiT2 History
431#431Claude 4.8 Opus Thinking Max EffortAnthropic50.51%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=claude-opus-4-8-max-effort25 Jun 2026livebench.aiT2 History
432#432gpt-5.2-2025-12-11-high50.25%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-5.2-2025-12-11-high25 Jun 2026livebench.aiT2 History
433#433Inkling xHigh Effort49.39%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=inkling-xhigh25 Jun 2026livebench.aiT2 History
434#434GPT-5.2-CodexOpenAI49.39%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-5.2-codex25 Jun 2026livebench.aiT2 History
435#435gemini-3.5-flash-high48.99%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gemini-3.5-flash-high25 Jun 2026livebench.aiT2 History
436#436Claude 4.6 Opus Thinking High EffortAnthropic48.99%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=claude-opus-4-6-thinking-auto-high-effort25 Jun 2026livebench.aiT2 History
437#437gpt-5.6-lunaOpenAI48.43%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-5.6-luna-max25 Jun 2026livebench.aiT2 History
438#438Kimi K2.6 ThinkingMoonshot AI46.92%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=kimi-k2.6-thinking25 Jun 2026livebench.aiT2 History
439#439DeepSeek V4 Flash (0731)DeepSeek46.77%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=deepseek-v4-flash-073125 Jun 2026livebench.aiT2 History
440#440gpt-5.4-nanoOpenAI46.77%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-5.4-nano-xhigh25 Jun 2026livebench.aiT2 History
441#441Grok Build 0.1xAI45.81%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=grok-build-0.125 Jun 2026livebench.aiT2 History
442#442Kimi K2.7 CodeMoonshot AI45.66%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=kimi-k2.7-code25 Jun 2026livebench.aiT2 History
443#443Gemini 3.5 Flash-Lite HighGoogle45.25%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gemini-3.5-flash-lite-high25 Jun 2026livebench.aiT2 History
444#444Gemini 3.1 Pro Preview HighGoogle44.14%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gemini-3.1-pro-preview-high25 Jun 2026livebench.aiT2 History
445#445Qwen3.7 MaxQwen43.59%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=qwen3.7-max25 Jun 2026livebench.aiT2 History
446#446Gemini 3.6 Flash HighGoogle43.43%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gemini-3.6-flash-high25 Jun 2026livebench.aiT2 History
447#447claude-sonnet-4-6-thinking-auto-medium-effort42.63%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=claude-sonnet-4-6-thinking-auto-medium-effort25 Jun 2026livebench.aiT2 History
448#448deepseek-v4-proDeepSeek42.63%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=deepseek-v4-pro25 Jun 2026livebench.aiT2 History
449#449gpt-5.4-miniOpenAI41.67%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=gpt-5.4-mini-xhigh25 Jun 2026livebench.aiT2 History
450#450Qwen3.6 PlusQwen41.36%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=qwen3.6-plus25 Jun 2026livebench.aiT2 History
451#451MiniMax M3MiniMax40.66%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=minimax-m325 Jun 2026livebench.aiT2 History
452#452claude-opus-4-5-20251101-thinking-64k-high-effort39.7%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=claude-opus-4-5-20251101-thinking-64k-high-effort25 Jun 2026livebench.aiT2 History
453#453Qwen3.6 27BQwen39.29%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=qwen3.6-27b25 Jun 2026livebench.aiT2 History
454#454Nemotron 3 UltraNVIDIA38.74%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=nemotron-3-ultra-550b-a55b25 Jun 2026livebench.aiT2 History
455#455deepseek-v4-flashDeepSeek37.63%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=deepseek-v4-flash25 Jun 2026livebench.aiT2 History
456#456Grok 4.3xAI18.54%release=2026-06-25 · subtasks=["javascript","typescript","python"] · livebench_model_id=grok-4.325 Jun 2026livebench.aiT2 History

Scores are reported as published, with their evaluation configuration (harness, prompting, judge). The bar is relative to the best score on this page. Results with different configs are not directly comparable — see methodology.

The config filter matches a value inside each result's configuration (server-side, `config=` on the API). Chips are the values shared by several rows on the first page; per-model identifiers are not offered.