InternVL2.5-8B-BoN-8 — benchmark results
Provider: Shanghai AI Lab. Released 2024-11-20. Access: Open.
Unified ELO 1513 ± 6, rank #785 of 1841 rated models, from 98 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LMM Reasoning - WeMath - Calculation of Solid Figures | 85.4 | Accuracy (%) | 93.8 |
| OpenVLM MathVista - VQA | 58.7 | Accuracy (%) | 87.9 |
| OpenVLM MathVista - SCI | 68.9 | Accuracy (%) | 84.2 |
| Open LMM Reasoning - WeMath - Three-step(S3) | 53.9 | Accuracy (%) | 84 |
| OpenVLM MathVista - STA | 80.4 | Accuracy (%) | 83.2 |
| OpenVLM MathVista - FQA | 71 | Accuracy (%) | 82.3 |
| OpenVLM MathVista - ARI | 64.6 | Accuracy (%) | 82 |
| OpenVLM MathVista - GPS | 73.6 | Accuracy (%) | 79.6 |
| Open LMM Reasoning - WeMath - Understanding of Solid Figures | 66.2 | Accuracy (%) | 79 |
| Open LMM Reasoning - DynaMath - Subject-graph theory | 14.6 | Accuracy (%) | 78.1 |
| Open LMM Reasoning - WeMath - Basic Transformations of Figures | 61.5 | Accuracy (%) | 76.5 |
| OpenVLM MathVista - GEO | 71.1 | Accuracy (%) | 76.5 |
Interactive version: theaggregate.ai/model?slug=internvl2-5-8b-bon-8 · How It Works · Data refreshed daily, snapshot 2026-07-25.