InternVL3.5-8B — benchmark results
OpenGVLab's 8B InternVL3.5 vision-language model (August 2025) with a Qwen3 backbone and Cascade RL training for multimodal reasoning. Provider: Shanghai AI Lab. Released 2025-08-26. Access: Open.
Unified ELO 1427 ± 25, rank #1095 of 1776 rated models, from 36 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AgroVG | 39.6 | All@.5 (self-reported) | 91.7 |
| CardioLens | 52.51 | F1 Score (Random) (self-reported) | 82.6 |
| EVALITA - faq | 53.86 | CPS | 72.3 |
| EVALITA - summarization-fanpage | 28.49 | CPS | 70.2 |
| MathVision | 52.05 | Overall Accuracy (%) | 62.7 |
| Math-VR | 40.8 | Overall Answer Correctness (self-reported) | 56.7 |
| MedLayXPlain | 58.7 | S (self-reported) | 54.8 |
| EVALITA - text-entailment | 74.32 | CPS | 51.1 |
| C3-Bench | 4.44 | Aggregation (self-reported) | 43.8 |
| EVALITA - word-in-context | 62.07 | CPS | 42.6 |
| VSI-Super-Wild | 32.18 | Overall Accuracy (%) | 41.7 |
| OCR-Robust | 57.26 | OCR1.0 Clean (self-reported) | 40 |
Interactive version: theaggregate.ai/model?slug=internvl3-5-8b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.