VeriSoftBench-Aristotle: leaderboard
Metric: Verification success rate (%) on the 100-task VeriSoftBench-Aristotle compatibility subset, full repository context plus the preceding same-file lemma statements, in a pre-compiled project environment (Aristotle runs as an end-to-end agent; Gemini-3-Pro with Pass@8 and three repair rounds); higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 2 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 65 | #77 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=verisoftbench-aristotle · How It Works · Data refreshed daily, snapshot 2026-10-11.