InternVL2.5-1B: benchmark results
Provider: Shanghai AI Lab. Released 2024-12-05. Access: Open.
Aggregate Rating 1368 ± 21, rank #7227 of 10916 rated models, from 165 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| FineBench - Object Manipulation | 79.6 | Accuracy (%; four-option multiple-choice questions on how a | 100 |
| OpenVLM OCRBench - Handwritten Mathematical Expression Recognition | 74 | Correct Answers (%) | 82.6 |
| OpenVLM MMMU - Diagnostics and Laboratory Medicine | 46.7 | Accuracy (%) | 82.3 |
| OpenVLM MMBench V1.1 EN - Action Recognition | 89.7 | Accuracy (%) | 78.1 |
| OpenVLM MMMU - Music | 33.3 | Accuracy (%) | 63.6 |
| OpenVLM MMMU - Chemistry | 46.7 | Accuracy (%) | 63.1 |
| OpenVLM MMBench V1.1 CN - Action Recognition | 86.3 | Accuracy (%) | 62.7 |
| OpenVLM MMBench V1.1 EN - Image Topic | 97.8 | Accuracy (%) | 62.4 |
| OpenVLM MMMU - Materials | 36.7 | Accuracy (%) | 61.5 |
| Cambrian-S Release - MVBench † | 64.3 | Accuracy (%) | 60 |
| OpenVLM OCRBench - Key Information Extraction | 79.5 | Correct Answers (%) | 57.9 |
| OpenVLM MMMU - Accounting | 43.3 | Accuracy (%) | 56.5 |
Interactive version: theaggregate.ai/model?slug=internvl2-5-1b · How It Works · Data refreshed daily, snapshot 2026-11-08.