MoNaCo: leaderboard
MoNaCo (AI2) — naturally complex questions whose answers require many-step aggregation across dozens to hundreds of Wikipedia pages.
Metric: F1. Source: tomerwolgithub.github.io. Saturation forecast: Around July 2027. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O3 | 61.18 | #124 |
| 2 | GPT-5 (Medium) | 60.11 | #98 (GPT-5) |
| 3 | Gemini 2.5 Pro | 59.11 | #152 |
| 4 | GPT-4o | 55.05 | #343 |
| 5 | DeepSeek V3 | 55.04 | #321 |
| 6 | Claude Opus 4 | 55.03 | #163 |
| 7 | O4 Mini | 54.92 | #180 |
| 8 | DeepSeek R1 | 53.82 | #251 |
| 9 | Gemini 2.5 Flash | 52.01 | #245 |
| 10 | O3 Mini | 48.75 | #272 |
| 11 | GPT-4 Turbo | 48.58 | #404 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=monaco · How It Works · Data refreshed daily, snapshot 2026-10-20.