AgentIdeaBench - Active: leaderboard
Metric: Weighted Hypothesis Score (1-10). Source: moyunxiang.com. 63 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 5 | 7.14 |
| 2 | GPT-5.6 Sol | 6.88 |
| 3 | GPT-5.6 Terra | 6.8 |
| 4 | Claude Opus 4.8 | 6.79 |
| 5 | Claude Opus 4.7 | 6.7 |
| 6 | GPT-5.5 | 6.68 |
| 7 | Claude Sonnet 5 | 6.66 |
| 8 | GPT-5.2 | 6.66 |
| 9 | GPT-5.4 | 6.62 |
| 10 | GPT-5.6 Luna | 6.43 |
| 11 | GPT-5 | 6.4 |
| 12 | GPT-5 Mini | 6.38 |
| 13 | GPT-5.4 Mini | 6.36 |
| 14 | GLM-5.1 | 6.33 |
| 15 | Gemini 3 Flash (Preview) | 6.29 |
Interactive version: theaggregate.ai/benchmark?slug=agentideabench-active · How It Works · Data refreshed daily, snapshot 2026-09-19.