InternVL3

An open multi-modal large-scale language model that excels in vision, reasoning and long text understanding, using multi-modal pre-training.

Introduction

Open MLLM excelling in vision, reasoning, and long context via multimodal pre-training.

What is InternVL?

InternVL is an open multi-modal large-scale language model (1B-78B) from OpenGVLab that performs well in vision, reasoning, long text understanding and agent capabilities, using native multi-modal pre-training. It surpasses basic language models on text tasks.

How to use InternVL?

You can ask InternVL questions. For example, you can ask someone what they are looking at, use Python to implement flowcharts, and make connections between different images.