New: Agent Index for knowledge work is live.Explore the index ↗

Best LLMs for reasoning with LiveBench

LiveBench: Reasoning

LiveBench 2.0 evaluates reasoning with LiveBench on its October 25, 2026 release. This suite contains 202 tasks or questions, scored from 0 to 100 in one run per model. Models use their highest supported effort unless a different setting is reported. This is an individual task-suite score, separate from LiveBench’s weighted overall average.

Updated: Nov 5, 2026Data as of: Nov 2, 2026

Widely usedNot saturated

Official benchmark

LiveBench: Reasoning results. Each row identifies its benchmark version and evaluation track. Scores from different versions are not directly comparable. Best effort from the latest stored report date per model and version.
Model / creator	Version	Reported score
Gemini 4 ArgonGoogle · Proprietary	2.0 / 2026-10-25	95.1%
GPT-6 AstraOpenAI · Proprietary	2.0 / 2026-10-25	92.7%
GPT-6.1 SolOpenAI · Proprietary	2.0 / 2026-10-25	92.6%
Claude Opus 5.5Anthropic · Proprietary	2.0 / 2026-10-25	92.2%
Claude Fable 5.1Anthropic · Proprietary	2.0 / 2026-10-25	91.7%
Claude Sonnet 5.5Anthropic · Proprietary	2.0 / 2026-10-25	91.6%
Kimi K3.1 ProMoonshot AI · Open source / open weights	2.0 / 2026-10-25	91.1%
Kimi K3Moonshot AI · Open source / open weights	2.0 / 2026-10-25	90.7%
GPT-5.6 TerraOpenAI · Proprietary	2.0 / 2026-10-25	90.6%
Muse Wonder 1.0Meta · Proprietary	2.0 / 2026-10-25	90.6%
Muse Spark 1.3Meta · Proprietary	2.0 / 2026-10-25	89.7%
Gemini 3.8 FlashGoogle · Proprietary	2.0 / 2026-10-25	89.3%
DeepSeek V4.1 Flash 1017DeepSeek · Open source / open weights	2.0 / 2026-10-25	89.1%
GPT-6.1 TerraOpenAI · Proprietary	2.0 / 2026-10-25	88.7%
Step 5 PreviewStepFun · Open source / open weights	2.0 / 2026-10-25	83.8%
MiniMax M3.1 FlashMiniMax · Open source / open weights	2.0 / 2026-10-25	82.9%
Grok 4.7xAI · Proprietary	2.0 / 2026-10-25	82.7%
Claude Haiku 5.5Anthropic · Proprietary	2.0 / 2026-10-25	82.4%
GPT-6 LunaOpenAI · Proprietary	2.0 / 2026-10-25	81.8%
Qwen3.8 27BQwen · Open source / open weights	2.0 / 2026-10-25	80%
MiMo-V2.6-ProXiaomi · Open source / open weights	2.0 / 2026-10-25	79.6%
InklingThinking Machines · Open source / open weights	2.0 / 2026-10-25	78.3%
GLM-5.3-FlashZ.ai · Open source / open weights	2.0 / 2026-10-25	77.6%
Nemotron 3.5 Super VLNVIDIA · Open source / open weights	2.0 / 2026-10-25	74.6%
Muse Spark OpenMeta · Open source / open weights	2.0 / 2026-10-25	73.7%

View all 25 published reports

Muse Wonder 1.090.6%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

GPT-5.6 Terra90.6%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

GPT-6.1 Terra88.7%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Step 5 Preview83.8%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Claude Haiku 5.582.4%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

GPT-6 Luna81.8%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

MiMo-V2.6-Pro79.6%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Nemotron 3.5 Super VL74.6%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Muse Spark Open73.7%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Gemini 4 Argon95.1%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

GPT-6 Astra92.7%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

GPT-6.1 Sol92.6%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Claude Opus 5.592.2%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Claude Fable 5.191.7%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Claude Sonnet 5.591.6%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Kimi K3.1 Pro91.1%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Muse Spark 1.389.7%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

Gemini 3.8 Flash89.3%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

DeepSeek V4.1 Flash 101789.1%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

MiniMax M3.1 Flash82.9%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning

GLM-5.3-Flash77.6%

2.0 / 2026-10-25

release: 2026-10-25 · Highest supported effort unless publisher reports an exception; single run · suite: reasoning