AgentIdeaBench - Active - Feasibility: leaderboard
Feasibility dimension of AgentIdeaBench's Active track, where the model gets a subfield name and a literature search tool and decides what to look up (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 dims field of reports/leaderboard_full.json; higher is better.
Metric: Feasibility Critic Score (1-10). Source: moyunxiang.com. Saturation forecast: Not forecast. 63 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 7 | #83 |
| 2 | GPT-5.2 | 7 | #115 |
| 3 | Llama 4 Maverick | 6.87 | #486 |
| 4 | GPT-5.4 Mini | 6.85 | #228 |
| 5 | Claude Opus 4.8 | 6.83 | #39 |
| 6 | Qwen 2.5 72B Instruct | 6.77 | #469 |
| 7 | Claude Opus 5 | 6.77 | #16 |
| 8 | GPT-5.6 Terra | 6.76 | #40 |
| 9 | GPT-5 Mini | 6.75 | #193 |
| 10 | GPT-5.6 Sol | 6.72 | #22 |
| 11 | Gemini 2.5 Flash Lite | 6.71 | #432 |
| 12 | Claude Opus 4.6 | 6.7 | #67 |
| 13 | GPT-4o Mini | 6.66 | #598 |
| 14 | GPT-5.4 Nano | 6.64 | #359 |
| 15 | GPT-5.5 | 6.63 | #33 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=agentideabench-active-feasibility · How It Works · Data refreshed daily, snapshot 2026-11-05.