Scientific General Intelligence on SGI-bench
37.74Deep ResearchInternAgent-1.5
Evaluation results
| Method | Links | ||
|---|---|---|---|
| InternAgent-1.5Base Model=Gemini-3-pro+o4-mini2026.02 | 37.74 | 58.11 | |
| Gemini-3-pro2026.02 | 18.48 | 39.68 | |
| Qwen3-Max2026.02 | 15.38 | 39.83 | |
| GPT-52026.02 | 14.47 | 55.4 | |
| Claude-Sonnet-4.52026.02 | 13.84 | 43.2 | |
| o4-mini2026.02 | 11.95 | 40.78 |