Skip to content

Benchmark

LiveBench javascript

Scores are percentages.

Models measured
51
Scores recorded
51
Sources
1
IQ weight
2.03

Score distribution

62.1%
Mean
63.6%
Median
10.9%
Std dev
54.5%
10th percentile
72.7%
90th percentile
18.2%
Min
18.2%77.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.03 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
1Muse Spark 1.1xhigh· LiveBench data77.3%
2Muse Spark 1.3xhigh· LiveBench data77.3%
3Qwen3.8 Max· LiveBench data77.3%
4Claude Opus 5.5xhigh· LiveBench data72.7%
5Gemini 3.8 Flashhigh· LiveBench data72.7%
6GLM 5.3· LiveBench data72.7%
7Grok 4.5· LiveBench data72.7%
8Grok 4.6· LiveBench data72.7%
9Muse Spark 1.2xhigh· LiveBench data72.7%
10Claude Sonnet 5xhigh· LiveBench data68.2%
11DeepSeek V4 Pro 0813· LiveBench data68.2%
12Gemini 3.7 Flashhigh· LiveBench data68.2%
13GPT 5.4xhigh· LiveBench data68.2%
14GPT-5.2 Codex· LiveBench data68.2%
15Inkling FP4xhigh· LiveBench data68.2%
16Kimi K3· LiveBench data68.2%
17Mistral Large 4high· LiveBench data68.2%
18Qwen 3.8 Flash Next· LiveBench data68.2%
19Claude Opus 4.6auto, high· LiveBench data63.6%
20Claude Opus 4.7xhigh· LiveBench data63.6%
21DeepSeek V4 Flash 0731· LiveBench data63.6%
22DeepSeek V4 Flash Vision Exp· LiveBench data63.6%
23Gemini 3.5 Flashhigh· LiveBench data63.6%
24Gemini 3.6 Flashhigh· LiveBench data63.6%
25GLM 5.2· LiveBench data63.6%
26GLM 5.3 Flash· LiveBench data63.6%
27GPT 5.4 Nanoxhigh· LiveBench data63.6%
28GPT 5.5xhigh· LiveBench data63.6%
29GPT-6.1 Solxhigh· LiveBench data63.6%
30Grok 4.7xhigh· LiveBench data63.6%
31Kimi K2.7 Code· LiveBench data63.6%
32MiniMax M3· LiveBench data63.6%
33Claude Haiku 5.5xhigh· LiveBench data59.1%
34Claude Opus 4.564k, high· LiveBench data59.1%
35Gemini 3.1 Pro Preview· LiveBench data59.1%
36Gemini 3.5 Flash Litehigh· LiveBench data59.1%
37GPT 5.2high· LiveBench data59.1%
38Grok Build 0.1· LiveBench data59.1%
39Kimi K2.6thinking· LiveBench data59.1%
40Qwen 3.8 27B· LiveBench data59.1%
41Qwen3.6 Plus· LiveBench data59.1%
42Qwen3.7 Max· LiveBench data59.1%
43Claude Sonnet 4.6auto, medium· LiveBench data54.5%
44DeepSeek V4 Flash· LiveBench data54.5%
45DeepSeek V4 Pro· LiveBench data54.5%
46Nemotron 3 Ultra· LiveBench data54.5%
47Qwen3.6 27B· LiveBench data54.5%
48GPT-5.4 minixhigh· LiveBench data50%
49Claude Sonnet 5.5xhigh· LiveBench data36.4%
50Grok 4.3· LiveBench data27.3%

Showing 1–50 of 51 results.