Question Answering on CORTEXBENCH Bridge
96.1AccuracyGPT-5.4
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-5.4Evaluation Setting=ORACLE2026.06 | 96.1 | |
| DeepSeek-V3.2Evaluation Setting=ORACLE2026.06 | 94.8 | |
| Claude Opus 4.6Evaluation Setting=ORACLE2026.06 | 91.9 | |
| DeepSeek-V4-FlashEvaluation Setting=ORACLE2026.06 | 90.2 | |
| Claude Sonnet 4.6Evaluation Setting=ORACLE2026.06 | 84 | |
| Gemini 3.1 ProEvaluation Setting=ORACLE2026.06 | 82.5 | |
| Gemini 3 ProEvaluation Setting=ORACLE2026.06 | 80.1 | |
| GPT-4oEvaluation Setting=ORACLE2026.06 | 77.5 | |
| Gemini 3.1 ProEvaluation Setting=1-SRCH2026.06 | 65.7 | |
| Gemini 3 ProEvaluation Setting=M-SRCH2026.06 | 63 | |
| Claude Opus 4.6Evaluation Setting=M-SRCH2026.06 | 62.5 | |
| GPT-5.4Evaluation Setting=M-SRCH2026.06 | 60.9 | |
| Claude Sonnet 4.6Evaluation Setting=M-SRCH2026.06 | 59.1 | |
| GPT-5.4Evaluation Setting=1-SRCH2026.06 | 58.4 | |
| DeepSeek-V4-FlashEvaluation Setting=M-SRCH2026.06 | 54.1 | |
| Gemini 3.1 ProEvaluation Setting=M-SRCH2026.06 | 52.5 | |
| Claude Sonnet 4.6Evaluation Setting=1-SRCH2026.06 | 52.5 | |
| DeepSeek-V3.2Evaluation Setting=M-SRCH2026.06 | 52.2 | |
| Claude Opus 4.6Evaluation Setting=1-SRCH2026.06 | 50.9 | |
| GPT-5.4Evaluation Setting=CLOSED2026.06 | 47.2 | |
| DeepSeek-V4-FlashEvaluation Setting=1-SRCH2026.06 | 44 | |
| DeepSeek-V3.2Evaluation Setting=1-SRCH2026.06 | 41.3 | |
| Gemini 3 ProEvaluation Setting=1-SRCH2026.06 | 38.1 | |
| Claude Opus 4.6Evaluation Setting=CLOSED2026.06 | 35.3 | |
| Claude Sonnet 4.6Evaluation Setting=CLOSED2026.06 | 34.7 | |
| GPT-4oEvaluation Setting=M-SRCH2026.06 | 33.8 | |
| DeepSeek-V4-FlashEvaluation Setting=CLOSED2026.06 | 30.9 | |
| Gemini 3 ProEvaluation Setting=CLOSED2026.06 | 30.9 | |
| DeepSeek-V3.2Evaluation Setting=CLOSED2026.06 | 30.3 | |
| Gemini 3.1 ProEvaluation Setting=CLOSED2026.06 | 28.1 | |
| GPT-4oEvaluation Setting=1-SRCH2026.06 | 26.9 | |
| GPT-4oEvaluation Setting=CLOSED2026.06 | 26.6 |