GPT-4o (0513, detail-high): benchmark results
Provider: OpenAI. Released 2024-05-13. Access: API.
Aggregate Rating 1615 ± 33, rank #884 of 10916 rated models, from 432 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM A-Bench - Generative Distortion | 72.4 | Accuracy (%) | 100 |
| OpenVLM Q-Bench - Yes-or-No (Others) | 82.3 | Accuracy (%) | 100 |
| OpenVLM MME - Commonsense Reasoning | 178.6 | Score | 99.8 |
| OpenVLM MMT-Bench - Color Recognition | 95 | Score (%) | 99.8 |
| OpenVLM LLaVA-Bench - Complex | 111.2 | Score | 99.6 |
| OpenVLM MMT-Bench - Animated Character Recognition | 90 | Score (%) | 99.5 |
| OpenVLM BLINK | 68 | Accuracy (%) | 99.4 |
| OpenVLM MMT-Bench - Human Interaction Understanding | 95 | Score (%) | 99.3 |
| OpenVLM MMT-Bench - OCR | 82.5 | Score (%) | 99.3 |
| OpenVLM MMT-Bench - Chemical Apparatus Recognition | 80 | Score (%) | 99 |
| OpenVLM MMT-Bench - Handwritten Retrieval | 90 | Score (%) | 99 |
| OpenVLM MMT-Bench - Health and Medicine | 92.9 | Score (%) | 99 |
Interactive version: theaggregate.ai/model?slug=gpt-4o-0513-detail-high · How It Works · Data refreshed daily, snapshot 2026-11-08.