Skip to content

GLM-4V-9B

An open multimodal model for bilingual visual dialogue

Visit official site

What GLM-4V-9B is

The model is described as a pretrained multimodal model supporting Chinese-English multi-turn dialogue and visual understanding at 1120x1120 resolution.

What you can do with it

  • Process images and generate descriptions
  • Hold multi-turn dialogue in two languages
  • Recognise text and interpret charts

Getting started

  • Import torch, PIL and transformers
  • Load the tokenizer from THUDM/glm-4v-9b
  • Prepare text and an RGB image, then apply the chat template

Pros & cons

✓ What we like

  • Python-based calls and text generation are documented
  • A model weight usage agreement is provided

! What to watch out for

  • The described call path needs torch, PIL and transformers
  • Images are converted to RGB before input

FAQ

What resolution is named?

The material states 1120x1120 high resolution.

How is it called?

The steps name Python libraries such as torch, PIL and transformers.

Last reviewed: 2026-11-15

More AI chatbot tools

View all →

How we review