T2 Theorem Testing: leaderboard
T2 (Theorem Testing, 2026) asks a model to write a Lean 4 theorem from its natural-language statement and repository context, then tests it like code: the theorem passes only if every successor theorem that depends on it still compiles. Testing accuracy (%) over 2,206 problems from five real-world Lean repositories; higher is better.
Metric: Testing accuracy (%): share of T2's 2,206 problems from five Lean 4 repositories in which the generated theorem, substituted for the original, lets every dependent successor theorem compile; zero-shot from a natural-language statement (written by Claude Sonnet 4.5) with predecessor and successor context, one completion at temperature 0.6, top-p 0.95; higher is better. Source: arxiv.org. Saturation forecast: Around 2030. 18 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.5 | 38.9 | #261 |
| 2 | GPT-5 Mini | 37.9 | #364 |
| 3 | GPT-5 | 37.7 | #178 |
| 4 | Llama 3.1 70B | 37 | #1499 |
| 5 | Claude 3.7 Sonnet | 36.8 | #470 |
| 6 | GPT-4o Mini | 36.7 | #1667 |
| 7 | GPT-5 Nano | 36.6 | #906 |
| 8 | DeepSeek R1 | 36.3 | #482 |
| 9 | Claude Sonnet 4 | 36 | #410 |
| 10 | Llama 3.1 405B | 33.2 | #1082 |
| 11 | GPT-OSS-120B | 32.3 | #704 |
| 12 | Goedel-Prover-V2-32B | 31 | |
| 13 | DeepSeek-Prover-V2-7B | 30 | |
| 14 | Llama 3.1 8B | 24.8 | #5196 |
| 15 | Goedel-Prover-V2-8B | 24.5 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=t2-theorem-testing · How It Works · Data refreshed daily, snapshot 2026-11-21.