LiveBench 2.0 Reasoning

Short logic, spatial and theory-of-mind questions with a single correct answer. LiveBench 2.0, refreshed with new questions each release to limit contamination.

CategoryReasoning
MetricScore (%)
SourceLiveBench (original)
LicenseNone stated
Benchmark pagelivebench.ai
StatusScored
Weight0.8 within Reasoning
FitDiscrimination 0.50, noise 0.34

Results

#ModelProviderResultDate
1 Gemini 4 Argon Google 95.1% 2026-11-06
2 GPT-6 Astra OpenAI 92.7% 2026-11-06
3 GPT-6.1 Sol OpenAI 92.6% 2026-11-06
4 Claude Opus 5.5 Anthropic 92.2% 2026-11-06
5 Claude Fable 5.1 Anthropic 91.7% 2026-11-06
6 Claude Sonnet 5.5 Anthropic 91.6% 2026-11-06
7 Grok 4.6 xAI 91.2% 2026-11-06
8 Kimi K3.1 Pro Moonshot AI 91.1% 2026-11-06
9 Kimi K3 Moonshot AI 90.7% 2026-11-06
10 Muse Wonder Meta 90.6% 2026-11-06
11 GPT-5.6 Terra OpenAI 90.6% 2026-11-06
12 Gemini 4 Carbon Google 90.6% 2026-11-06
13 Muse Spark 1.3 Meta 89.7% 2026-11-06
14 Gemini 3.8 Flash Google 89.3% 2026-11-06
15 DeepSeek V4.1 Flash 1017 DeepSeek 89.1% 2026-11-06
16 Claude Fable 5.2 Anthropic 88.8% 2026-11-06
17 GPT-6.1 Terra OpenAI 88.7% 2026-11-06
18 Qwen3.8 Max Alibaba 88.2% 2026-11-06
19 Mistral Large 4 Preview Mistral 85.2% 2026-11-06
20 Mistral Large 4 Mistral 83.9% 2026-11-06
21 Step 5 Preview StepFun 83.8% 2026-11-06
22 MiniMax-M3.1-Flash MiniMax 82.9% 2026-11-06
23 Grok 4.7 xAI 82.7% 2026-11-06
24 Claude Haiku 5.5 Anthropic 82.4% 2026-11-06
25 GPT-6 Luna OpenAI 81.8% 2026-11-06
26 Qwen3.8 27B Alibaba 80.0% 2026-11-06
27 MiMo-V2.6-Pro Xiaomi 79.6% 2026-11-06
28 Inkling Thinking Machines 78.3% 2026-11-06
29 GLM-5.3-Flash Zhipu AI 77.6% 2026-11-06
30 Nemotron 3.5 Super VL 120B A12B NVIDIA 74.6% 2026-11-06
31 Muse Spark Open Meta 73.7% 2026-11-06
32 Gemma 4 31B IT Google 68.3% 2026-11-06