METR data deduplication
Agents write an efficient algorithm that removes near-duplicate lines from a large file. A trial succeeds when F1 reaches at least 80%.
Anthropic reports the share of trials scoring above the F1 threshold. The card calls it a METR public task.
- Level
- Expert Specialists with years of training, often over hours of work
- 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.
Claude Opus 4 leads with 32.6% among models released by May 2025
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 May 2025, and nobody has reported a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6 Astra
- Claude Opus 5.5
- GPT-6.1 Terra
- GPT-6.1 Sol
- and 16 more
- 1 Claude Opus 4 32.6%
- 2 Claude Sonnet 4 27.6%
- 3 Claude 3.7 Sonnet 13.3%
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 4 | 32.6% | May 2025 | Model card | |
| 2 | Claude Sonnet 4 | 27.6% | May 2025 | Model card | |
| 3 | Claude 3.7 Sonnet | 13.3% | Feb 2025 | Model card |
2 models have held number one since 2023
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.
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 |
|---|---|---|---|---|
| 22 May 2025 | Claude Sonnet 4 system card | 27.6 | METR data deduplication, 8/29 trials above the F1 80% threshold (27.6%). Median F1 74.9% (7.3.2) | AI R&D risk |
| 22 May 2025 | Claude Opus 4 system card | 32.6 | METR data deduplication, 15/46 trials above the F1 80% threshold (32.6%). Median F1 76.2% (7.3.2) | AI R&D risk |
| 24 Feb 2025 | Claude 3.7 Sonnet system card | 13.3 | METR Data Deduplication, 4/30 trials above the F1 80% threshold. Median F1 70.2% (7.2.2, Figure 20) | 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%