Skip to content

Benchmark

LiveBench integrals_with_game

Scores are percentages.

Models measured
51
Scores recorded
51
Sources
1
IQ weight
1.80

Score distribution

73%
Mean
78%
Median
19.3%
Std dev
45%
10th percentile
96%
90th percentile
24%
Min
24%100%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.80 for chat models, measured over 51 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude Sonnet 5.5xhigh· LiveBench data100%
2Claude Opus 5.5xhigh· LiveBench data99%
3GPT 5.5xhigh· LiveBench data99%
4GPT-6.1 Solxhigh· LiveBench data99%
5Grok 4.7xhigh· LiveBench data96%
6Muse Spark 1.3xhigh· LiveBench data96%
7DeepSeek V4 Pro 0813· LiveBench data94%
8GPT 5.4xhigh· LiveBench data93%
9GPT 5.2high· LiveBench data92%
10Muse Spark 1.2xhigh· LiveBench data91%
11Claude Sonnet 4.6auto, medium· LiveBench data90%
12Mistral Large 4high· LiveBench data90%
13Claude Haiku 5.5xhigh· LiveBench data89%
14Claude Sonnet 5xhigh· LiveBench data88%
15Gemini 3.7 Flashhigh· LiveBench data88%
16Grok 4.6· LiveBench data85%
17Claude Opus 4.7xhigh· LiveBench data84%
18Muse Spark 1.1xhigh· LiveBench data83%
19GPT 5.4 Nanoxhigh· LiveBench data82.8%
20Qwen3.8 Max· LiveBench data81%
21Gemini 3.8 Flashhigh· LiveBench data80%
22Nemotron 3 Ultra· LiveBench data80%
23Claude Opus 4.564k, high· LiveBench data78%
24DeepSeek V4 Pro· LiveBench data78%
25Gemini 3.1 Pro Preview· LiveBench data78%
26Grok 4.5· LiveBench data78%
27DeepSeek V4 Flash Vision Exp· LiveBench data76%
28GLM 5.2· LiveBench data76%
29GPT-5.2 Codex· LiveBench data75%
30Claude Opus 4.6auto, high· LiveBench data73%
31Inkling FP4xhigh· LiveBench data72%
32GLM 5.3· LiveBench data69%
33Gemini 3.5 Flashhigh· LiveBench data68%
34Grok Build 0.1· LiveBench data66%
35DeepSeek V4 Flash 0731· LiveBench data65%
36Qwen 3.8 27B· LiveBench data65%
37Gemini 3.6 Flashhigh· LiveBench data63%
38Grok 4.3· LiveBench data63%
39Qwen3.6 Plus· LiveBench data62%
40Qwen 3.8 Flash Next· LiveBench data59%
41Qwen3.7 Max· LiveBench data59%
42MiniMax M3· LiveBench data56%
43Kimi K2.6thinking· LiveBench data54%
44Kimi K3· LiveBench data54%
45Qwen3.6 27B· LiveBench data52%
46GLM 5.3 Flash· LiveBench data45%
47DeepSeek V4 Flash· LiveBench data37%
48GPT-5.4 minixhigh· LiveBench data37%
49Kimi K2.7 Code· LiveBench data37%
50Gemini 3.5 Flash Litehigh· LiveBench data26%

Showing 1–50 of 51 results.