GPT-4v (1106, detail-low): benchmark results
Provider: OpenAI. Released 2023-11-06. Access: API.
Aggregate Rating 1498 ± 33, rank #2384 of 10916 rated models, from 204 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM SEEDBench IMG - Instance Interaction | 81.4 | Accuracy (%) | 98.4 |
| OpenVLM MMMU - Psychology | 80 | Accuracy (%) | 95.4 |
| OpenVLM MME - Posters | 187.8 | Score | 94 |
| OpenVLM LLaVA-Bench - Complex | 102.3 | Score | 91.8 |
| OpenVLM LLaVA-Bench - Detail | 90.1 | Score | 91.8 |
| OpenVLM AI2D - Faults and Earthquakes | 78.6 | Accuracy (%) | 90.1 |
| OpenVLM MMMU - Art Theory | 90 | Accuracy (%) | 89.9 |
| LLaVA-Bench | 93.1 | Score | 88.2 |
| OpenVLM MME - OCR | 177.5 | Score | 85 |
| OpenVLM MMMU - Accounting | 60 | Accuracy (%) | 83 |
| OpenVLM MME - Commonsense Reasoning | 155 | Score | 82.5 |
| OpenVLM MMBench V1.1 EN - Image Emotion | 81.1 | Accuracy (%) | 82.3 |
Interactive version: theaggregate.ai/model?slug=gpt-4v-1106-detail-low · How It Works · Data refreshed daily, snapshot 2026-11-08.