EO-Gym: leaderboard
Earth Observation (EO) analysis is inherently interactive: resolving uncertainty often requires expanding the region of interest, retrieving historical observations, and switching across sensors such as optical and Synthetic Aperture Radar. However, most EO benchmarks collapse this process into fixed-input, single-turn tasks. To address this gap, we present EO-Gym, a controlled executable framewor
Metric: Pass@3 (self-reported). Source: benchmarklist.com. Saturation forecast: Around February 2027. 11 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 67 | #152 |
| 2 | GPT-4.1 Mini | 64 | #355 |
| 3 | GPT-4.1 | 64 | #248 |
| 4 | Gemini 2.5 Flash | 63 | #245 |
| 5 | GPT-5.1 | 63 | #135 |
| 6 | Qwen 3 VL 32B Instruct | 59 | #283 |
| 7 | Qwen 3 VL 8B (Thinking) | 52 | |
| 8 | Qwen 3 VL 8B Instruct | 49 | #420 |
| 9 | Qwen 3 VL 4B Instruct | 49 | #515 |
| 10 | Qwen 3 VL 4B (Thinking) | 49 | #480 (Qwen 3 VL 4B) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=eo-gym · How It Works · Data refreshed daily, snapshot 2026-10-20.