Skip to content

LLaVA-o1

A vision language model that reasons step by step

Visit official site

What LLaVA-o1 is

The model is described as a visual language model that reasons step by step across image and language inputs.

What you can do with it

  • Analyse a problem in staged reasoning
  • Combine image and language information

Getting started

  • Download the code and pretrained weights
  • Set up the runtime with required libraries

Pros & cons

✓ What we like

  • An arXiv paper supports the work
  • Image and language inputs are handled together

! What to watch out for

  • A runtime with libraries must be set up
  • Pretrained weights have to be loaded first

FAQ

Who developed it?

The material names a Peking University team.

What is released?

Open source code and pretrained weights are listed.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review