FeynmanBench - Momentum Routing: leaderboard
Metric: CP4 pass rate (%), a self-consistent internal momentum assignment conserving momentum at every vertex, judged by a three-model LLM panel, on FeynmanBench's Feynman-diagram tasks (a rendered diagram plus minimal conventions; Standard Model QED, electroweak and QCD diagrams from tree level to two loops), API models at temperature 0, mean of five runs; four models ran on subsets of 165 to 742 items, the others on about 2,000; higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 19 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 23.9 |
| 2 | Claude Sonnet 4.6 | 12.6 |
| 3 | Qwen 3.6 Plus | 9.7 |
| 4 | Qwen 3.5 Plus | 9.4 |
| 5 | GLM-5V Turbo | 7 |
| 6 | Seed 2.0 Pro | 5.4 |
| 7 | Gemini 3 Flash | 5 |
| 8 | Claude Opus 4.7 | 4.7 |
| 9 | Claude Opus 4.6 | 4.5 |
| 10 | GPT-5.4 | 2.1 |
| 11 | GPT-5.1 | 1.3 |
| 12 | GPT-5.2 | 0.9 |
| 13 | Claude Opus 4.5 | 0.9 |
| 14 | Claude Sonnet 4.5 | 0.6 |
| 15 | Qwen 3.6 Max | 0.5 |
Interactive version: theaggregate.ai/benchmark?slug=feynmanbench-momentum-routing · How It Works · Data refreshed daily, snapshot 2026-10-07.