Remote Sensing Reasoning on XLRS-Bench
60.4PASS@1Baseline + pre-warming + ES Text QA (w/ CoT)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Baseline + pre-warming + ES Text QA (w/ CoT)Parameters=7B, Pre-warming=ES Text QA, Chain-of-Thought (CoT)=true2026.02 | 60.4 | 96.25 | — | |
| Intern-S1Parameters=241B2026.02 | 55 | — | — | |
| Baseline + pre-warming with SuperRS-VQAParameters=7B, Pre-warming=SuperRS-VQA2026.02 | 52.39 | 91.85 | — | |
| Intern-S1-miniParameters=8B2026.02 | 51.6 | — | — | |
| GeoLLaVA-8KParameters=7B2026.02 | 51.5 | — | — | |
| Qwen3-VL-235B-A22BParameters=235B2026.02 | 51.11 | — | — | |
| Qwen3-VL-8BParameters=8B2026.02 | 50.02 | — | — | |
| BaselineParameters=7B2026.02 | 50.01 | 82.58 | — | |
| GLM-4.1VParameters=9B2026.02 | 49.8 | — | — | |
| GPT-5.22026.02 | 47.53 | — | — | |
| Qwen2.5-VLParameters=7B2026.02 | 47.4 | 85.75 | — | |
| Grok-42026.02 | 45.4 | — | — | |
| Gemini 2.5 Pro2026.02 | 45.2 | — | — | |
| GPT-o32026.02 | 43.6 | — | — | |
| Claude 3.7 Sonnet2026.02 | 40.5 | — | — | |
| ZoomEarthParameters=3B2026.02 | 40.2 | — | — | |
| VLM-R^3Parameters=7B, Tools=true2026.02 | 39.1 | — | — | |
| GeoChatParameters=7B2026.02 | 22.03 | — | — | |
| BREENParameters=7B2026.05 | — | — | 37.02 | |
| Coarse-to-fineParameters=7B2026.05 | — | — | 35.65 | |
| DeepSeek-VL2Parameters=27B2026.05 | — | — | 38.31 | |
| Emu3Parameters=8B2026.05 | — | — | 32.53 | |
| EVEv2Parameters=7B2026.05 | — | — | 38.41 | |
| GeoChatParameters=7B2026.05 | — | — | 26.56 | |
| GeoLLaVA-8KParameters=7B2026.05 | — | — | 35.89 | |
| SkyNativeParameters=8B2026.05 | — | — | 39.62 | |
| VHMParameters=7B2026.05 | — | — | 28.25 | |
| ZoomEarthParameters=3B2026.05 | — | — | 30.65 |