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
51Claude Sonnet 4.6max· Epoch AI Benchmarking Hub data76.3%
52Gemini 3.1 Pro Previewhigh· Epoch AI Benchmarking Hub data76%
53Qwen3.6 Flash· Epoch AI Benchmarking Hub data76%
54DeepSeek V4.1 Flash· Epoch AI Benchmarking Hub data75%
55Gemini 3.6 Flashhigh· Epoch AI Benchmarking Hub data75%
56GPT 5.5xhigh· Epoch AI Benchmarking Hub data75%
57GPT-5.6 Solmax· Epoch AI Benchmarking Hub data75%
58Kimi K3.1 Promax· Epoch AI Benchmarking Hub data75%
59Muse Wonder 1.0· Epoch AI Benchmarking Hub data75%
60Qwen Plus· Epoch AI Benchmarking Hub data75%
61Qwen3.7 Plus· Epoch AI Benchmarking Hub data75%
62Claude Sonnet 4· Epoch AI Benchmarking Hub data74.9%
63Gemini 3.7 Flashhigh· Epoch AI Benchmarking Hub data74.7%
64Gemma 4 31B IT· Epoch AI Benchmarking Hub data74.7%
65Grok 4 Fast· Epoch AI Benchmarking Hub data74.7%
66Qwen3.5 Flash· Epoch AI Benchmarking Hub data74.6%
67DeepSeek V3.1 Terminus· Epoch AI Benchmarking Hub data74.3%
68Gemini 4 Argonhigh· Epoch AI Benchmarking Hub data73.3%
69Kimi K2.6· Epoch AI Benchmarking Hub data73.3%
70Nemotron 3 Ultra· Epoch AI Benchmarking Hub data73.3%
71Qwen3.6 35B A3B· Epoch AI Benchmarking Hub data73.3%
72GPT 4o· Epoch AI Benchmarking Hub data73.2%
73DeepSeek V3.1thinking· Epoch AI Benchmarking Hub data72.9%
74GPT-5 minihigh· Epoch AI Benchmarking Hub data72.4%
75Qwen3 30B A3B Instruct· Epoch AI Benchmarking Hub data72.4%
76Qwen3.5 9B· Epoch AI Benchmarking Hub data72.4%
77GPT 4.1· Epoch AI Benchmarking Hub data71.6%
78DeepSeek V3· Epoch AI Benchmarking Hub data71.2%
79o4 Minihigh· Epoch AI Benchmarking Hub data71.1%
80DeepSeek V3.2 Expthinking· Epoch AI Benchmarking Hub data70.7%
81Claude Haiku 4.5· Epoch AI Benchmarking Hub data70.4%
82Claude Haiku 5.5max· Epoch AI Benchmarking Hub data70.4%
83Gemini 3.1 Flash Litehigh· Epoch AI Benchmarking Hub data70.3%
84DeepSeek V4 Promax· Epoch AI Benchmarking Hub data69.7%
85Gemini 3.5 Flash Litehigh· Epoch AI Benchmarking Hub data69.7%
86GPT-5.6 Terramax· Epoch AI Benchmarking Hub data69.7%
87GPT-5.6 Lunamax· Epoch AI Benchmarking Hub data69.3%
88Qwen3.8 Maxxhigh· Epoch AI Benchmarking Hub data69.3%
89Claude 3.5 Sonnet· Epoch AI Benchmarking Hub data69%
90Gemma 4 26B A4B· Epoch AI Benchmarking Hub data68.9%
91Qwen3 32B· Epoch AI Benchmarking Hub data68.5%
92DeepSeek V4 Pro 0813· Epoch AI Benchmarking Hub data68.4%
93Gemini 2.5 Flash· Epoch AI Benchmarking Hub data68.4%
94Kimi K3max· Epoch AI Benchmarking Hub data68.4%
95MiMo V2.5 Pro· Epoch AI Benchmarking Hub data68.4%
96Qwen3 8B· Epoch AI Benchmarking Hub data68.4%
97Mistral Large 4· Epoch AI Benchmarking Hub data68.1%
98DeepSeek Chat· Epoch AI Benchmarking Hub data67.1%
99Gemini 3.8 Flash· Epoch AI Benchmarking Hub data67.1%
100Gemini 4 Carbon· Epoch AI Benchmarking Hub data67.1%

Showing 51–100 of 177 results.

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