Skip to content

Benchmark

LiveBench connections

Scores are percentages.

Models measured
51
Scores recorded
51
Sources
1
IQ weight
2.47

Score distribution

93.6%
Mean
97.3%
Median
11.7%
Std dev
89.3%
10th percentile
100%
90th percentile
34%
Min
34%100%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.47 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
1DeepSeek V4 Pro 0813· LiveBench data100%
2Gemini 3.1 Pro Preview· LiveBench data100%
3Gemini 3.5 Flashhigh· LiveBench data100%
4Gemini 3.6 Flashhigh· LiveBench data100%
5Gemini 3.7 Flashhigh· LiveBench data100%
6Gemini 3.8 Flashhigh· LiveBench data100%
7GPT 5.4xhigh· LiveBench data100%
8GPT 5.5xhigh· LiveBench data100%
9GPT-6.1 Solxhigh· LiveBench data100%
10Grok 4.6· LiveBench data100%
11Kimi K3· LiveBench data100%
12Muse Spark 1.2xhigh· LiveBench data100%
13Claude Opus 4.564k, high· LiveBench data99.3%
14Claude Opus 4.6auto, high· LiveBench data99.3%
15Claude Opus 5.5xhigh· LiveBench data99.3%
16Claude Sonnet 4.6auto, medium· LiveBench data99.3%
17DeepSeek V4 Flash Vision Exp· LiveBench data99.3%
18GPT 5.2high· LiveBench data99%
19GPT-5.4 minixhigh· LiveBench data99%
20Claude Sonnet 5.5xhigh· LiveBench data98.5%
21DeepSeek V4 Pro· LiveBench data98%
22Grok 4.5· LiveBench data98%
23Grok 4.7xhigh· LiveBench data98%
24Kimi K2.7 Code· LiveBench data98%
25Muse Spark 1.1xhigh· LiveBench data98%
26DeepSeek V4 Flash 0731· LiveBench data97.3%
27Qwen3.7 Max· LiveBench data96.5%
28MiniMax M3· LiveBench data96.3%
29Claude Sonnet 5xhigh· LiveBench data95.5%
30GLM 5.3· LiveBench data95.3%
31GPT-5.2 Codex· LiveBench data95%
32Qwen3.8 Max· LiveBench data94.5%
33GPT 5.4 Nanoxhigh· LiveBench data94.4%
34GLM 5.2· LiveBench data94%
35Grok 4.3· LiveBench data94%
36Muse Spark 1.3xhigh· LiveBench data94%
37Claude Haiku 5.5xhigh· LiveBench data93.8%
38GLM 5.3 Flash· LiveBench data93.7%
39Claude Opus 4.7xhigh· LiveBench data92.3%
40Gemini 3.5 Flash Litehigh· LiveBench data92%
41Inkling FP4xhigh· LiveBench data91.2%
42Qwen 3.8 27B· LiveBench data90.5%
43Qwen3.6 Plus· LiveBench data90.5%
44Qwen 3.8 Flash Next· LiveBench data90.2%
45DeepSeek V4 Flash· LiveBench data89.5%
46Kimi K2.6thinking· LiveBench data89.3%
47Nemotron 3 Ultra· LiveBench data86.7%
48Grok Build 0.1· LiveBench data84%
49Qwen3.6 27B· LiveBench data77.2%
50Mistral Large 4high· LiveBench data48%

Showing 1–50 of 51 results.