Core Capabilities
When you begin a conversation with le Chat, you're engaging with a powerful generative AI model developed by Mistral AI.
You can communicate using natural language to ask questions or assign tasks. Employing prompt engineering techniques can enhance the results you receive.
Beyond its core capabilities, le Chat can utilize various tools to accomplish tasks. It selects the appropriate tools based on the context of your conversation and the features you have enabled.
Canvas is an interface designed for collaborating with le Chat on writing and coding projects that require editing and revisions.
With Canvas, le Chat gains a deeper understanding of your goals. You can highlight specific sections to direct le Chat's focus precisely.
You can edit text or code directly and use shortcuts to have le Chat adjust writing length, debug code, and perform other useful actions quickly.
Canvas supports React/HTML rendering in a sandbox environment, allowing the use of all npm packages and many JavaScript libraries.
The language models behind le Chat are trained on a vast amount of global information, but their knowledge is limited to the data available at the time of training. As a result, le Chat may not have up-to-date information on specialized topics or current events. By enabling web search, le Chat can retrieve additional information from the internet to better answer questions and complete tasks.
When you upload an image as part of your conversation, le Chat uses Mistral AI’s vision model to interpret the image. This enables le Chat to answer questions about the image, or use information in the image as context for your queries.
When you upload text documents (PDF, txt…), le Chat uses tools to extract text and find relevant information. This helps le Chat understand the contents of the document and supports tasks like summarization, Q&A…
Le Chat can utilize a specific tool called Code Interpreter, a dedicated code sandbox environment designed to handle data documents (such as Excel, CSV, JSON) and address math-related questions. With Code Interpreter, le Chat can answer quantitative questions about your data, correct common data errors, and generate data visualizations by executing Python code.
Le Chat's image generation feature allows users to create visual content based on textual descriptions or prompts. This capability is powered by Black Forest Labs Flux Pro and Ultra models that can interpret natural language inputs to produce relevant and creative images.