Mistral AI has just announced the biggest number of changes to its generative AI work assistant ‘le Chat,’ while detailing the all-new model Pixtral Large.
The Paris-based AI startup has just leveled up its offerings, with its latest announcement seeing the introduction of the new Pixtral Large. This multimodal model is said to be “significantly more powerful” than the previous version and it has now been integrated into le Chat.
The upgrade is a 124B open-weights model built on top of Mistral Large 2, making it the second model in the multimodal family. It has a 128K context window, fitting a minimum of 30 high-resolution images.
It’s not just the new model that the team has been working on though, there are new functions for the work assistant on Monday (November 18) which will be available in beta.
This includes web searchability, with the tool now having training knowledge with internet search which aims to help users when it comes to learning, researching, and working on the web.
Another upgrade is Canvas which is a new interface that will pop up in the le Chat window when people “need to go beyond conversations and into ideation.” It’s a collaboration tool where teams can create documents, presentations, code, mockups, and more.
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
Faster document understanding, le Chat Agents and image generation
Le Chat is also described as now having faster and more accurate document understanding as it can “process large, complex PDF documents and images, for analysis and summarization.”
This is due to Pixtral Large which is working in conjunction with an experimental model that the company says will be explained in a few weeks.
Through the document understanding ability, users can: “Simply attach documents to your chat bar and get insights on anything—graphs, tables, diagrams, text, formulae, equations… basically, the entirety of the document.”
The startup has also partnered up with Black Forest Labs to bring their image generation models to le Chat, with this now able to be completed within the AI assistant.
Finally, the latest change includes le Chat Agents to automate repetitive workflows.
“If you find yourself reusing prompts, or if you’ve identified certain patterns of creative work that you prefer to be automated, you can codify these workflows as agents. Popular use cases include receipt scanning for expense reporting, creating TL;DR’s of long meeting minutes, and invoice processing.”
These agents can be published and shared with teammates, with all of these features now being offered in beta for free.
Featured Image: Mistral AI Pixtral Large