InternVL2-8B: benchmark results
OpenGVLab's 8B vision-language model (July 2024) pairing InternViT-300M with internlm2_5-7b-chat for image, document and video understanding. Provider: Shanghai AI Lab. Released 2024-07-15. Access: Open.
Unified ELO 1461 ± 1, rank #892 of 1392 rated models, from 1043 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LMM Reasoning - MMMath - Difficulty-easy | 45.4 | Accuracy (%) | 100 |
| Open LMM Reasoning - MMMath - Knowledge-L1-Properties of Shapes | 29.4 | Accuracy (%) | 100 |
| Open LMM Reasoning - MMMath - Knowledge-L2-Acute Angle Trigonometric Functions | 19.7 | Accuracy (%) | 100 |
| Open LMM Reasoning - MMMath - Knowledge-L2-Circle | 34.3 | Accuracy (%) | 100 |
| Open LMM Reasoning - MMMath - Knowledge-L2-Intersecting and Parallel Lines | 33.8 | Accuracy (%) | 100 |
| Open LMM Reasoning - MMMath - Knowledge-L2-Symmetry of Shapes | 11.9 | Accuracy (%) | 100 |
| OpenVLM MMT-Bench - Disease Diagnosis | 100 | Score (%) | 99.8 |
| OpenVLM MMT-Bench - Chemical Apparatus Recognition | 80 | Score (%) | 99 |
| MEGA-Bench Task - Code Solution Compare | 50 | Task Score (%) | 98.8 |
| MEGA-Bench Task - TQA Textbook QA | 100 | Task Score (%) | 97.7 |
| OpenVLM MMT-Bench - GUI Install | 65 | Score (%) | 97.1 |
| OpenVLM MMBench V1.1 CN - Image Scene | 92.4 | Accuracy (%) | 96.6 |
Interactive version: theaggregate.ai/model?slug=internvl2-8b · How It Works · Data refreshed daily, snapshot 2026-09-05.