Novel compiler task (Anthropic)
Claude Opus 5 80.9% leads since Jul 2026 Only in system cards
Agents build a compiler for an unusual programming language from a specification and test cases.
In-house Anthropic evaluation. The tasks are not public.
Made by Anthropic Released May 2025 Machine learning engineering Tasks kept private Last in a system card Jul 2026
Last model added Jul 2026
- 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
- Code A program or patch run against tests.
4
Models scored in lab cards
11
System cards report it, from 1 lab
Claude Opus 5 leads with 80.9% among models released by Jul 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 Jul 2026, 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 15 more
- 1 Claude Opus 5 80.9%
- 2 Claude Mythos Preview 77.2%
- 3 Claude Sonnet 5 76.1%
- 4 Claude Opus 4.7 71.1%
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Claude Opus 5 | 80.9% | Jul 2026 | Model card | |
| 2 | Claude Mythos Preview | 77.2% | Apr 2026 | Model card | |
| 3 | Claude Sonnet 5 | 76.1% | Jun 2026 | Model card | |
| 4 | Claude Opus 4.7 | 71.1% | Apr 2026 | Model card |
2 models have held number one since its release in May 2025
Lab claims count from the day the card came out, independent runs from the model’s release.
Anthropic
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Reported by the lab in its model card
Best so far
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.
11 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 |
|---|---|---|---|---|
| 24 Jul 2026 | Claude Opus 5 system card | 80.9 | Novel Compiler, pass rate on complex tests (Table 2.3.5.A) | AI R&D risk |
| 30 Jun 2026 | Claude Sonnet 5 system card | 76.1 | Novel Compiler, pass rate on complex tests 76.1% (Table 2.3.3.A) | AI R&D risk |
| 16 Apr 2026 | Claude Opus 4.7 system card | 71.1 | Novel Compiler, pass rate on complex tests (Table 2.3.4.A) | AI R&D risk |
| 7 Apr 2026 | Claude Mythos Preview system card | 77.2 | Novel Compiler, pass rate on complex tests 77.2% (Table 2.3.3.A) | AI R&D risk |
| 17 Feb 2026 | Claude Sonnet 4.6 system card | Novel compiler, 93.7 of basic tests and 67.27 of complex tests passed, no overall (6.3.3.6) | AI R&D risk | |
| 5 Feb 2026 | Claude Opus 4.6 system card | Novel compiler, two test sets with no overall: 98.2% of basic tests and 65.83% of complex tests passed, threshold 90% on complex tests (8.3.3.6) | AI R&D risk | |
| 24 Nov 2025 | Claude Opus 4.5 system card | Novel compiler, two test sets with no overall: 93.7% of basic tests and 69.37% of complex tests passed, threshold 90% on complex tests (7.3.2.6) | AI R&D risk | |
| 29 Sep 2025 | Claude Sonnet 4.5 system card | Novel compiler, two test sets with no overall: mean pass rate 81.7% on basic tests and 29.7% on advanced tests, threshold 90% (9.3.3.6) | AI R&D risk | |
| 5 Aug 2025 | Claude Opus 4.1 system card addendum | Novel compiler, no overall: mean pass rate 74.4% on basic tests and 6.81% on advanced tests (6.4.1) | AI R&D risk | |
| 22 May 2025 | Claude Sonnet 4 system card | Novel compiler, no overall: mean pass rate 50.00% on basic tests and 17.08% on advanced tests. The 7.3 summary says 52.96% on basic tests (7.3.3.6) | AI R&D risk | |
| 22 May 2025 | Claude Opus 4 system card | Novel compiler, no overall: mean pass rate 64.44% on basic tests and 9.44% on advanced tests (7.3.3.6) | 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%