Skip to content

Benchmark

ZEROBench

Scores are percentages. Higher is better.

Models measured
84
Scores recorded
90
Sources
3
IQ weight
2.09

Score distribution

5.4%
Mean
2%
Median
7.8%
Std dev
0%
10th percentile
13.4%
90th percentile
0%
Min
0%43.2%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.09 for chat models, measured over 84 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-6 Astramax· ZeroBench43.2%
2Claude Opus 5.5max· ZeroBench35%
3GPT-6 Solmax· ZeroBench28.8%
4GPT-5.6 Solmax· ZeroBench22%
5Claude Opus 5max· ZeroBench17.2%
6GPT 5.5xhigh· ZeroBench17.2%
7GPT 5.4xhigh· ZeroBench16.4%
8Claude Fable 5max· ZeroBench15.8%
9GPT-5.6 Terramax· ZeroBench14.2%
10GPT-5.6 Lunamax· ZeroBench13.4%
11Gemini 3.5 Flash· ZeroBench12.8%
12Gemini 3.6 Flashhigh· ZeroBench12.2%
13GPT-6 Lunamax· ZeroBench12.2%
14Qwen3.5 397B A17BQwen3.5 release/model card · published setting· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.12%
15Gemini 3.1 Pro Preview· ZeroBench11.6%
16Grok 4.5high· ZeroBench11.4%
17GPT 5.2medium reasoning· ZeroBench11.2%
18Doubao Seed 1.8· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim11%
19Grok 4.6xhigh· ZeroBench11%
20Claude Opus 4.8· ZeroBench10.4%
21Gemini 3 Pro Preview· ZeroBench10%
22Claude Sonnet 5max· ZeroBench9.4%
23Gemini 3 Flash· ZeroBench9.4%
24GPT-5.4 mini· ZeroBench9%
25Claude Opus 4.7xhigh· ZeroBench8.8%
26Claude Sonnet 4.6· ZeroBench7.8%
27Claude Opus 4.6· ZeroBench6.4%
28Grok 4.3high· ZeroBench6.4%
29GPT-5 minihigh reasoning· ZeroBench6.2%
30Grok 4.20non-reasoning· ZeroBench6.2%
31Gemini 3.5 Flash Litehigh· ZeroBench6%
32Grok 4.20reasoning· ZeroBench6%
33Qwen3.5 Omni PlusQwen3.5-Omni release · published setting· Reported by Qwen, Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.5%
34o4 Mini· ZeroBench4.6%
35Claude Opus 4.5· ZeroBench3.6%
36Claude Opus 4.1thinking· ZeroBench3.4%
37Gemini 3.1 Flash Lite· ZeroBench3.4%
38Claude Sonnet 4thinking· ZeroBench2.8%
39Gemini 2.5 Pro· ZeroBench2.8%
40Claude Opus 4thinking· ZeroBench2.4%
41Claude Sonnet 3.7thinking· ZeroBench2.4%
42OpenAI o3· ZeroBench2.4%
43Claude Sonnet 4· ZeroBench2.2%
44GPT 5medium reasoning· ZeroBench2.2%
45Claude Sonnet 4.5thinking· ZeroBench2%
46GPT-4.5 Preview· ZeroBench2%
47Claude Opus 4.1· ZeroBench1.8%
48GPT 5.1medium reasoning· ZeroBench1.8%
49Llama 4 Scout· ZeroBench1.6%
50Gemini 2.5 Flash· ZeroBench1.4%

Showing 1–50 of 90 results.

Data credits: Data by LLM Stats · BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.