Skip to content

Pixtral-Large-Instruct-2411

A 124B multimodal model from Mistral AI

Visit official site

What Pixtral-Large-Instruct-2411 is

The model is described as a 124B-parameter multimodal system from Mistral AI built on Mistral Large 2.

What you can do with it

  • Understand documents, charts and natural images
  • Keep text understanding from Mistral Large 2
  • Handle a 128K context window

Model notes

  • A 123B multimodal decoder is listed
  • A 1B vision encoder is listed
  • The context fits at least 30 high-resolution images

Pros & cons

✓ What we like

  • Text performance carries over from Mistral Large 2
  • A standard instruction template V7 is listed

! What to watch out for

  • Serving is described through vLLM
  • Requests need text and image URLs

FAQ

How large is it?

The material states 124B parameters, with a 123B multimodal decoder and a 1B vision encoder.

Which datasets are named?

MathVista, DocVQA and VQAv2 are listed.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review