Model Details
InternVL-Chat-V1.5
Quick answer: InternVL-Chat-V1.5 is OpenGVLab's 20B-parameter open-weight multimodal model, notable for strong OCR and high-resolution image understanding, and one of the top-scoring open models on the MME benchmark at release.
At a Glance
Where InternVL-Chat-V1.5 leads
- Strong OCR and high-resolution image handling among open-weight models at release
Where it lags
- Superseded by later InternVL2/InternVL2.5 releases on most current benchmarks
Best for: open-weight comparison baseline in multimodal benchmark literature.
What InternVL-Chat-V1.5 Is
InternVL-Chat-V1.5 is part of OpenGVLab's InternVL series of open-weight multimodal LLMs, built with a large vision encoder paired with a 20B LLM backbone, emphasizing dynamic high-resolution image processing.
Specifications
| Field | Value |
|---|---|
| Organization | OpenGVLab (Shanghai AI Laboratory) |
| Modality | Multimodal (text + vision) |
| License | Open weights (research license) |
| Release date | April 2024 |