TransCore-M: benchmark results
Provider: PCI Research. Released 2024-02-13.
Aggregate Rating 1325 ± 37, rank #8247 of 10916 rated models, from 438 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| OpenVLM A-Bench - Nouns as Adjectives | 86 | Accuracy (%) | 97.2 |
| OpenVLM MMT-Bench - 3D Indoor Recognition | 50 | Score (%) | 96.4 |
| OpenVLM CCBench - Sketch Reasoning | 93.3 | Accuracy (%) | 95.2 |
| OpenVLM SEEDBench IMG - Text Understanding | 77.4 | Accuracy (%) | 95.1 |
| OpenVLM MMT-Bench - 3D | 55 | Score (%) | 90.8 |
| OpenVLM MMT-Bench - Face Mask Anomaly Detection | 65 | Score (%) | 90.8 |
| OpenVLM CCBench - Calligraphy and Painting | 78.9 | Accuracy (%) | 89.4 |
| OpenVLM CCBench - Traditional Shows | 71.2 | Accuracy (%) | 87.4 |
| OpenVLM COCO Captions | 33.5 | CIDEr | 87.1 |
| OpenVLM CCBench - Cultural Relics | 69.1 | Accuracy (%) | 87 |
| OpenVLM MMBench V1.0 EN - AR | 85.8 | Accuracy (%) | 86.7 |
| OpenVLM CCBench - Food and Clothes | 65.2 | Accuracy (%) | 86.3 |
Interactive version: theaggregate.ai/model?slug=transcore-m · How It Works · Data refreshed daily, snapshot 2026-11-08.