Monorepo-Bench
Agents replicate pull-request style contributions in a large internal OpenAI repository. Graded by hidden unit tests as pass@1.
In-house OpenAI AI self-improvement evaluation. The tasks are not public.
- Level
- Research frontier Researchers, on open problems or tasks that take days
- Scored
- Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
- Answers
- Agent task Many steps in a terminal, browser, desktop, app, game or robot.
GPT-6 Astra leads with 76% among models released by Sep 2026
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
The newest model with a score here came out in Sep 2026, and nobody has reported a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- Claude Opus 5.5
- GPT-6.1 Terra
- GPT-6.1 Sol
- Claude Sonnet 5.5
- and 14 more
- 1 GPT-6 Astra 76%
- 2 GPT-5.5 60%
- 3 GPT-5.4 59.3%
- 4 GPT-5.3 Codex 56%
- 5 GPT-5.4 Mini 54%
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-6 Astra | 76% | Sep 2026 | Model card | |
| 2 | GPT-5.5 | 60% | Apr 2026 | Model card | |
| 3 | GPT-5.4 | 59.3% | Mar 2026 | Model card | |
| 4 | GPT-5.3 Codex | 56% | Feb 2026 | Model card | |
| 5 | GPT-5.4 Mini | 54% | Mar 2026 | Model card |
4 models have held number one since its release in Feb 2026
Lab claims count from the day the card came out, independent runs from the model’s release.
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Labs report it only in system cards
We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark. Some system cards do, below.
5 system cards report it, from 1 lab
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| 3 Sep 2026 | GPT-6 Astra system card | 76 | OAI-repoBench (Monorepo-Bench) native unit-test pass rate, baseline condition, 50 tasks, medium reasoning effort, printed chart label (Figure 37) | Monitorability |
| 23 Apr 2026 | GPT-5.5 system card | 60 | Monorepo-Bench, pass@1, printed chart label (Figure 18, 9.1.3.1) | AI R&D risk |
| 17 Mar 2026 | GPT-5.4 Mini system card | 54 | Monorepo-Bench, pass@1 (Table 20) | AI R&D risk |
| 5 Mar 2026 | GPT-5.4 system card | 59.3 | Monorepo-Bench, pass@1, printed chart label (Figure 15, 5.1.3.1) | AI R&D risk |
| 5 Feb 2026 | GPT-5.3 Codex system card | 56 | Monorepo-Bench, pass@1, printed chart label (Figure 8, 5.1.3.1) | AI R&D risk |
More in AI research
- MLS-Bench In use Agent must invent one algorithmic change in an ML research scaffold whose gain transfers across settings. Scored relative to human baselines. Claude Fable 5.1 50.3%
- PaperBench In use Agent replicates ICML 2024 papers from scratch. LLM judge grades against author-written hierarchical rubrics.
Qwen 3.8 Max 93% - MLE-bench In use Kaggle-style ML engineering competitions. Agent trains models and is scored against competition leaderboards. GPT-6 Astra 93.8%
- PostTrainBench 1.2 In use CLI agent gets a base model, one H100 and a time budget to post-train it for a target benchmark. Scored on normalized gain. Claude Fable 5.1 44.6%
- KernelBench In use Replace PyTorch operators with faster custom GPU kernels. Scored on correctness and speedup.
MiniMax-M3 28.8% - OpenAI PRs Only in system cards Agents replicate internal OpenAI pull requests from a pre-PR branch and a prompt. Graded by hidden unit tests as pass@1. GPT-5.2 55%