French artificial intelligence startup Mistral AI recently made a major update to its AI assistant, Le Chat, which significantly improved its capabilities and performance. The update covers web search, image generation and the launch of the new Pixtral Large model, aiming to provide users with a stronger and more convenient AI experience. The upgrade of Le Chat not only enhances information acquisition capabilities, but also improves content creation and editing efficiency, marking the continuous progress and innovation of Mistral AI in the field of AI technology.
French artificial intelligence startup Mistral AI has announced the addition of a range of new features to its Le Chat AI assistant, including integrated web search, image generation and the newly launched Pixtral Large model.
Le Chat Feature Upgrade
Users can now directly access real-time web content through Le Chat to easily obtain the information they need. At the same time, with the help of Black Forest Labs' Flux Pro model, users can also generate high-quality images to meet a variety of creative needs.
In addition to web search and image generation, Le Chat also introduces a canvas interface that allows users to edit generated content directly within the chat window. This feature enables users to write documents, create presentations, and edit code without regenerating responses, greatly improving work efficiency.
Launch the Pixtral Large Model
The Pixtral Large model launched by Mistral AI performs very well in visual tasks. The model is built on Mistral Large2 and has achieved outstanding results in multiple industry benchmarks.
For example, in the MathVista mathematical reasoning test, the Pixtral Large score reached 69.4%, surpassing other competitors such as GPT-4o and Gemini1.5Pro.
At the same time, the model's ability to analyze charts and complex documents has been recognized, and it can process a variety of information including graphics, tables, and formulas.
The Pixtral Large model combines a multimodal decoder with a 123 billion parameter and a visual encoder with a 1 billion parameter, which can process up to 128 high-resolution images simultaneously, with a maximum context window of 30K.
This makes it excellent in document analysis and complex image processing. Mistral AI said that Pixtral Large will also provide academic and commercial licenses on the Hugging Face platform, which will facilitate research and application by different users.
In addition, Mistral AI has updated its Mistral Large language model to improve long context understanding and the accuracy of function calls.
The updated model will be available through the Mistral API and will be available soon on Google Cloud and Microsoft Azure.
Pixtral Large Paper Portal: https://arxiv.org/abs/2410.07073
Model page: https://huggingface.co/mistralai/Pixtral-Large-Instruct-2411
Le Chat entrance: https://auth.mistral.ai/ui/login?flow=b3e9d399-afc8-497b-8f8d-99900b447c08
API portal: https://docs.mistral.ai/api/
Key points:
Mistral AI adds web search and image generation functions to Le Chat Assistant, and users can obtain web content in real time.
The new canvas interface makes document writing, demonstration production and code editing more convenient.
The Pixtral Large model outperforms the performance of its major competitors in multiple visual tasks.
In short, this update of Mistral AI not only enhances the practicality and efficiency of Le Chat, but also demonstrates its leading technical strength in the fields of large language models and multimodal models. The launch of the Pixtral Large model further consolidates Mistral AI's competitive position in the field of artificial intelligence. In the future, the development of Mistral AI deserves continuous attention.