Mistral AI Unveils Pixtral Large: A 124B Parameter Open-Source Multimodal Model, Elevating le Chat

Explore Pixtral Large, Mistral AI's 124B open-source multimodal model. Upgrade le Chat with real-time search, PDF processing, and image generation.

Waking up to another big move from Mistral AI!

Today, Mistral AI's multimodal family welcomes a new heavyweight member: the Pixtral Large foundational model.

Built upon Mistral Large 2, this multimodal model features 124 billion open weights, boasting cutting-edge image comprehension capabilities. It can interpret documents, charts, and natural images while retaining Mistral Large 2’s superior text-understanding abilities.

In addition to this new model release, Mistral AI has also upgraded its free chatbot, le Chat, adding image generation, web search, and interactive canvas features—directly challenging ChatGPT.

All these features are available as a free beta.

Each Mistral AI update excites the entire AI community.

As someone remarked, “Six months ago, the gap between open-source and closed-source models was enormous. Now, the most advanced AI is quickly becoming accessible to anyone who wants to use it.”

Just how powerful is it? Let’s dive into the details.