Skip to content

Benchmark

LiveBench Language

Scores are percentages.

Models measured
42
Scores recorded
46
Sources
1
IQ weight
2.05

Score distribution

39.5%
Mean
39.4%
Median
14.9%
Std dev
23.3%
10th percentile
59.9%
90th percentile
9.2%
Min
9.2%80.2%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.05 for chat models, measured over 41 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1GPT 5.1high· Epoch AI Benchmarking Hub data80.2%
2Gemini 2.5 Pro· Epoch AI Benchmarking Hub data67.8%
3o1high· Epoch AI Benchmarking Hub data65.4%
4GPT-4.5 Preview· Epoch AI Benchmarking Hub data61.5%
5Claude Sonnet 3.7· Epoch AI Benchmarking Hub data59.9%
6Qwen2.5-Max· Epoch AI Benchmarking Hub data56.3%
7Claude 3.5 Sonnet· Epoch AI Benchmarking Hub data53.8%
8QwQ 32B· Epoch AI Benchmarking Hub data51.4%
9gemini-exp· Epoch AI Benchmarking Hub data51.3%
10o3 Minihigh· Epoch AI Benchmarking Hub data50.7%
11Claude 3 Opus· Epoch AI Benchmarking Hub data50.4%
12DeepSeek V3· Epoch AI Benchmarking Hub data49.1%
13DeepSeek R1· Epoch AI Benchmarking Hub data48.5%
14GPT 4o· Epoch AI Benchmarking Hub data47.6%
15o3 Minimedium· Epoch AI Benchmarking Hub data46.3%
16Grok 2· Epoch AI Benchmarking Hub data45.6%
17gemini-2.0-pro-exp-02-05· Epoch AI Benchmarking Hub data44.9%
18Perplexity Sonar· Epoch AI Benchmarking Hub data44.1%
19Gemini 2.0 Flash Experimentalthinking· Epoch AI Benchmarking Hub data42.2%
20Learnlm 1.5 Pro Experimental· Epoch AI Benchmarking Hub data42%
21O1 Minimedium· Epoch AI Benchmarking Hub data40.9%
22Gemini 2.0 Flash· Epoch AI Benchmarking Hub data40.7%
23Mistral Large2411· Epoch AI Benchmarking Hub data39.4%
24Llama 3.3 70B Instruct· Epoch AI Benchmarking Hub data39.2%
25o3 Minilow· Epoch AI Benchmarking Hub data38.3%
26Gemini 2.0 Flash Experimental· Epoch AI Benchmarking Hub data38.2%
27Nova Prov1· Epoch AI Benchmarking Hub data37%
28Claude Haiku 3.5· Epoch AI Benchmarking Hub data35.4%
29Gemma 3 27B IT· Epoch AI Benchmarking Hub data34.6%
30Gemini 2.0 Flash-Lite Preview· Epoch AI Benchmarking Hub data34.3%
31Gemini 2.0 Flash Lite· Epoch AI Benchmarking Hub data33.6%
32Gemma 2 27B IT· Epoch AI Benchmarking Hub data32.6%
33Mistral Small· Epoch AI Benchmarking Hub data30.5%
34Mistral Small2503· Epoch AI Benchmarking Hub data29.1%
35GPT-4o mini· Epoch AI Benchmarking Hub data28.6%
36DeepSeek R1 Distill Qwen 32B· Epoch AI Benchmarking Hub data26.8%
37Nova Litev1· Epoch AI Benchmarking Hub data25.9%
38Phi 4· Epoch AI Benchmarking Hub data25.6%
39Gemma 2 9B IT· Epoch AI Benchmarking Hub data25.5%
40DeepSeek R1 Distill Llama 70b· Epoch AI Benchmarking Hub data23.8%
41Qwen 2.5 Coder 32B Instruct· Epoch AI Benchmarking Hub data23.3%
42QwQ 32B Preview· Epoch AI Benchmarking Hub data21.1%
43C4AI Command R 08 2024· Epoch AI Benchmarking Hub data16.7%
44Nova Microv1· Epoch AI Benchmarking Hub data15.8%
45Phi-3 Small 8K Instruct· Epoch AI Benchmarking Hub data12.9%
46Phi-3 Mini 4K Instruct· Epoch AI Benchmarking Hub data9.2%

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence)