Google RE-Bench (GRB)
Gemini 3.7 Flash 27% Only in system cards Google DeepMind internal research-engineering benchmark with a 2-hour timeout per task, used for the ML R&D CCL rule-out.
About a quarter of original tasks had bugs. Argon was rescored on the remaining subset.
- 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.
Gemini 3.7 Flash leads with 27%
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.
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Gemini 3.7 Flash | 27% | Aug 2026 | Model card |
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.
3 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 |
|---|---|---|---|---|
| 13 Oct 2026 | Gemini 4 Argon system card | Google RE-Bench (GRB), pass@1 on subset of original tasks: no point estimate reported because of infrastructure timeouts; roughly half of completed attempts solved, below 90% CCL threshold (3.4.1.3) | AI R&D risk | |
| 13 Oct 2026 | Gemini 4 Argon model card | ML R&D acceleration level 1: internal benchmark of complex research engineering tasks, below CCL threshold, no score (Frontier Safety table, p.13) | AI R&D risk | |
| 13 Aug 2026 | Gemini 3.7 Flash system card | 27 | GRB internal benchmark, 74 tasks, average pass@1. About 20% of tasks bugged, adjusted maximum 47% (Results, p.37) | 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%