Multi-Docker-Eval: leaderboard
Tests LLMs on constructing executable Docker environments for real-world software repositories. Given a GitHub repo, the model must produce a working Dockerfile and configuration.
Metric: Resolved (%). Source: github.com. Saturation forecast: Around June 2027. 16 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K2.5 | 41.82 | #144 |
| 2 | GLM-4.7 | 40.62 | #192 |
| 3 | DeepSeek V3.1 | 37.72 | #264 |
| 4 | Kimi K2 0905 | 37.62 | #288 |
| 5 | DeepSeek V3.2 | 36.63 | #205 |
| 6 | Kimi K2 (Thinking) | 36.53 | #243 (Kimi K2) |
| 7 | Claude Sonnet 4 | 35.53 | #199 |
| 8 | Kimi K2 (0711) | 34.23 | #492 |
| 9 | GPT-5 Mini | 34.13 | #186 |
| 10 | MiMo-V2-Flash | 32.83 | #372 |
| 11 | Gemini 2.5 Flash | 29.44 | #245 |
| 12 | GPT-OSS-120B | 27.35 | #341 |
| 13 | DeepSeek R1 | 26.65 | #251 |
| 14 | Qwen 3 235B A22B | 23.65 | #313 |
| 15 | GPT-OSS-20B | 17.17 | #518 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=multi-docker-eval · How It Works · Data refreshed daily, snapshot 2026-10-20.