Skip to content

Benchmark

Dtbench EDT answer preference

Coding benchmarks compare models on programming tasks. Scores are percentages.

Models measured
170
Scores recorded
177
Sources
1
IQ weight
0.99

Score distribution

67.8%
Mean
69%
Median
11.3%
Std dev
53%
10th percentile
81.6%
90th percentile
43.3%
Min
43.3%96.1%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.99 for chat models, measured over 170 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 Fable 5.2· Epoch AI Benchmarking Hub data96.1%
2Claude Opus 5.5max· Epoch AI Benchmarking Hub data92.1%
3Claude Sonnet 5.5max· Epoch AI Benchmarking Hub data89.5%
4Claude Fable 5max· Epoch AI Benchmarking Hub data87.4%
5Claude Opus 4.7max· Epoch AI Benchmarking Hub data87.1%
6Claude Opus 5max· Epoch AI Benchmarking Hub data86.8%
7GPT 5high· Epoch AI Benchmarking Hub data85.3%
8Claude Opus 4.8max· Epoch AI Benchmarking Hub data84.8%
9Claude Opus 4.6max· Epoch AI Benchmarking Hub data84.3%
10Grok 4.5high· Epoch AI Benchmarking Hub data84%
11OpenAI o3high· Epoch AI Benchmarking Hub data84%
12Claude Fable 5.1· Epoch AI Benchmarking Hub data82.9%
13Grok 4.20· Epoch AI Benchmarking Hub data82.7%
14Grok 4.6xhigh· Epoch AI Benchmarking Hub data82.7%
15Grok 4.7xhigh· Epoch AI Benchmarking Hub data82.7%
16Qwen3.6 27B· Epoch AI Benchmarking Hub data82.4%
17Claude Opus 4.5high· Epoch AI Benchmarking Hub data82.1%
18GPT-5.5 Proxhigh· Epoch AI Benchmarking Hub data81.6%
19GPT-6 Astra· Epoch AI Benchmarking Hub data81.6%
20O3 Prohigh· Epoch AI Benchmarking Hub data81.3%
21Qwen 3 235B A22B Thinking 2507· Epoch AI Benchmarking Hub data81.3%
22Gemini 2.5 Pro· Epoch AI Benchmarking Hub data81.1%
23GPT 5.2xhigh· Epoch AI Benchmarking Hub data81.1%
24Qwen3.5 122B-A10Bnone· Epoch AI Benchmarking Hub data80.8%
25Claude Sonnet 5max· Epoch AI Benchmarking Hub data80%
26Qwen3.5 Plus· Epoch AI Benchmarking Hub data79.7%
27Claude Opus 4· Epoch AI Benchmarking Hub data79.5%
28Claude Sonnet 4.5· Epoch AI Benchmarking Hub data79.5%
29Claude Opus 4.1· Epoch AI Benchmarking Hub data79.3%
30Qwen 3 235B A22B· Epoch AI Benchmarking Hub data78.7%
31Qwen3 235B A22B Instruct 2507· Epoch AI Benchmarking Hub data78.7%
32Qwen3.5 27B· Epoch AI Benchmarking Hub data78.7%
33Qwen3.5 397B A17B· Epoch AI Benchmarking Hub data78.7%
34Qwen3.6 Plus· Epoch AI Benchmarking Hub data78.7%
35Gemini 3.5 Flashhigh· Epoch AI Benchmarking Hub data77.7%
36GLM 5.2max· Epoch AI Benchmarking Hub data77.6%
37GPT 5.1high· Epoch AI Benchmarking Hub data77.6%
38GPT-6 Solmax· Epoch AI Benchmarking Hub data77.6%
39Qwen3.6 Max Preview· Epoch AI Benchmarking Hub data77.6%
40Qwen3.5 35B A3B· Epoch AI Benchmarking Hub data77.5%
41GLM 5.3· Epoch AI Benchmarking Hub data77.3%
42GPT-6.1 Solmax· Epoch AI Benchmarking Hub data77.3%
43Grok 4.1 Fastreasoning· Epoch AI Benchmarking Hub data77.3%
44Grok 4.3high· Epoch AI Benchmarking Hub data77.3%
45Qwen3 Max· Epoch AI Benchmarking Hub data77.3%
46Qwen3.7 Max· Epoch AI Benchmarking Hub data77%
47Gemini 3 Flashhigh· Epoch AI Benchmarking Hub data76.5%
48GPT-4.1 mini· Epoch AI Benchmarking Hub data76.4%
49DeepSeek V4 Flashmax· Epoch AI Benchmarking Hub data76.3%
50GPT-5.6 Solpro / max· Epoch AI Benchmarking Hub data76.3%

Showing 1–50 of 177 results.

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