Interactive Agent Task on TRIP-Bench Overall
45Loose Success ScoreGPT-5.2
Evaluation results
| Method | Links | ||
|---|---|---|---|
| GPT-5.2thinking mode=w/ thinking2026.02 | 45 | 18.5 | |
| DeepSeek-V3.2thinking mode=w/ thinking2026.02 | 40 | 10.5 | |
| Claude-Sonnet-4.5thinking mode=w/ thinking2026.02 | 32 | 8.5 | |
| Gemini-3-Flashthinking mode=w/ thinking2026.02 | 23.3 | 6.3 | |
| GLM-4.7thinking mode=w/ thinking2026.02 | 20.3 | 4 | |
| Gemini-3-Prothinking mode=w/ thinking2026.02 | 20 | 2.8 | |
| DeepSeek-V3.2thinking mode=w/o thinking2026.02 | 18.5 | 2.3 | |
| Gemini-3-Prothinking mode=w/o thinking2026.02 | 18 | 3 | |
| Claude-Sonnet-4.5thinking mode=w/o thinking2026.02 | 17.3 | 1.8 | |
| Gemini-3-Flashthinking mode=w/o thinking2026.02 | 17.3 | 5.5 | |
| GLM-4.7thinking mode=w/o thinking2026.02 | 14.8 | 0 | |
| GPT-5.2thinking mode=w/o thinking2026.02 | 13.3 | 0.5 | |
| Kimi-K2-Thinkingthinking mode=w/ thinking2026.02 | 10.8 | 2.3 | |
| Qwen3-235B-A22B-Instruct-2507thinking mode=w/o thinking2026.02 | 5.8 | 0.5 | |
| Kimi-K2-0905-Previewthinking mode=w/o thinking2026.02 | 3.3 | 0 | |
| Qwen3-235B-A22B-Thinking-2507thinking mode=w/ thinking2026.02 | 0 | 0 |