DoGBench: leaderboard

Agents write or update user-facing software documentation in response to real repository events in open-source projects such as Helm and PostHog, including cases where the right answer is to change nothing. Patches are graded against rubrics validated with project maintainers and combined into a 0-100 composite. Each row is one model in one coding harness. Held-out split, higher is better. Source: dogbench.ai.

Metric: Composite score (0-100), held-out split (117 items). Source: dogbench.ai. Saturation forecast: Around July 2028. 7 models tracked. Updated 3 Oct 2026.

Top models

#ModelScoreOverall rank
1Qwen 3.8 Max47.26#64
2GPT-5.6 Sol46.22#33
3GLM-5.244.16#152
4Kimi K2.7 Code43.88#199
5Claude Opus 4.841.16#62
6Claude Sonnet 4.640.02#173
7GPT-5.534.02#50

No result here: #1 Claude Fable 5.2, #2 GPT-6.2 Astra, #3 Gemini 4 Carbon.

Interactive version: theaggregate.ai/benchmark?slug=dogbench · How It Works · Data refreshed daily, snapshot 2026-11-17.