Multimodal model

Pixtral 12B

Pixtral 12B is Mistral’s mid-sized vision–language model. It accepts images plus text prompts and generates grounded answers in text, with solid OCR, chart/diagram reasoning, and screenshot/UI understanding. It supports long-context input, function/tool calling, and JSON outputs, balancing quality and efficiency.

Type
Multimodal
Weights
Open
Downloads
155k
Likes
106
Pipeline
image text to text
License
apache-2.0

Model hub

mistral-experimental/pixtral-12b

Repo
mistral-experimental/pixtral-12b
Library
transformers
Access
Public
Technical tags and languages
transformers safetensors llava image-text-to-text conversational endpoints_compatible
Open on Hugging Face

Related tools

1

More from Mistral AI

12

About Mistral AI

1

Mistral AI develops frontier generative AI models, infrastructure, and enterprise deployment products.

Benchmark scores

7

How Pixtral 12B ranks on tracked AI benchmarks. Click any benchmark to see its full leaderboard.

Tools using Pixtral 12B

12