AgentIdeaBench - Static - Originality: leaderboard
Originality dimension of AgentIdeaBench's Static track, where the model is handed the references for a subfield (arXiv 2609.07611v1). Mean LLM-critic score on a 1-10 scale, dropping the highest of three critics and averaging over ideas and subfields, from the static_dims field of reports/leaderboard_full.json; higher is better.
Metric: Originality Critic Score (1-10). Source: moyunxiang.com. Saturation forecast: Around November 2027. 63 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5 | 7.25 | #16 |
| 2 | GPT-5.6 Sol | 7.13 | #22 |
| 3 | GPT-5.5 | 7.08 | #33 |
| 4 | GPT-5.6 Terra | 7.03 | #40 |
| 5 | GPT-5.6 Luna | 6.81 | #79 |
| 6 | GPT-5 | 6.67 | #105 |
| 7 | GPT-5.4 | 6.49 | #83 |
| 8 | GPT-5.2 | 6.42 | #115 |
| 9 | GPT-5.4 Mini | 6.4 | #228 |
| 10 | Claude Sonnet 5 | 6.21 | #65 |
| 11 | Claude Opus 4.7 | 6.14 | #51 |
| 12 | Gemini 3.5 Flash | 6.14 | #66 |
| 13 | Claude Opus 4.8 | 6.1 | #39 |
| 14 | GPT-5 Mini | 6.1 | #193 |
| 15 | GPT-5.1 | 6 | #145 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=agentideabench-static-originality · How It Works · Data refreshed daily, snapshot 2026-11-05.