Mistral Enhances Le Chat with Internet Surfing and Image Generation

French AI startup Mistral has unveiled a series of updates to its product line, including the new Pixtral Large model and the addition of image generation to Le Chat.

We’re proud to introduce the next generation of le Chat. Search, PDF upload, coding, image generation, le Canevas… All in one place: https://t.co/0P2NfFqULn https://t.co/qYTHs6vXX9

— Mistral AI (@MistralAI) November 18, 2024

The chatbot has been taught to perform internet searches with quotes in the search bar, similar to OpenAI’s ChatGPT.

Developers have added a “canvas” tool, enabling users to modify or edit content, such as webpage layouts.

“You can use [the canvas] for creating documents, presentations, code, layouts. […] Modify content without regenerating responses, draft and preview your projects,” the company’s blog states.

Additionally, Le Chat can process large PDF documents, images, files with graphs and equations for analysis and summarization. The AI can automate workflows such as scanning expense reports and processing invoices.

The latest feature is made possible by the new multimodal AI model Pixtral Large, which is “significantly more powerful” than the previous Pixtral 12B. It boasts 124 billion parameters.

Image generation in Le Chat has been added through a partnership with Black Forest Labs. Previously, Elon Musk’s xAI integrated a similar tool into the Grok-2 chatbot following a similar collaboration.

In October, Mistral released its first generative artificial intelligence models designed for use on edge devices like laptops and phones.

Earlier, the startup introduced its first multimodal model capable of processing images and text.

In June, the company raised $640 million in a Series B funding round, valuing it at $6 billion.