GPT-4V: benchmark results
Provider: OpenAI. Access: API.
Aggregate Rating 1561 ± 18, rank #1540 of 10916 rated models, from 62 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OmniACT | 17.02 | Action Score | 100 |
| REBUS | 24 | Accuracy | 100 |
| CMMMU | 43.7 | Test Overall (%) | 96.9 |
| ChartBench | 50.74 | Five-task average, Acc+ and NumberQA (%) | 95 |
| ChartBench - Chart Recognition | 96.19 | Acc+ (%) | 95 |
| GeoMMBench - Applications | 69.5 | Micro-averaged accuracy (%) on the remote sensing applicatio | 94.1 |
| CMMMU (Validation) | 42.5 | Validation Overall (%) | 93.8 |
| MMNeedle (1 Image, 4*4 Stitching, Exact Accuracy) | 54.72 | 1 Image, 4*4 Stitching, Exact Accuracy | 90.9 |
| MMNeedle (10 Images, 4*4 Stitching, Exact Accuracy) | 7.58 | 10 Images, 4*4 Stitching, Exact Accuracy | 90.9 |
| OmniStarPro-RNG (Offline) | 4.97 | Semantic correctness (0-10; GPT-4o judge score of each narra | 90.9 |
| OmniStarPro-RNG (Offline) - Summary Fluency | 5.37 | Summary fluency (0-10; GPT-4o judge rating of the concatenat | 90.9 |
| ChartBench - Global Conception | 46.19 | Acc+ (%) | 90 |
Interactive version: theaggregate.ai/model?slug=gpt-4v · How It Works · Data refreshed daily, snapshot 2026-11-08.