Mistral
Introduction
Mistral is an AI service that gives AAU researchers access to generative AI tools. In this section you can read more about the two services: Le Chat for conversational AI assistance and AI Studio for larger generative tasks, including API-based integration with custom AI workflows.
Getting Started
How to access
Learn how access to Mistral is expected to work at AAU
Le Chat
Start with conversational AI for drafting, ideation, and exploration
AI Studio
Work more systematically with prompts, models, and reusable setups
Key Features
Two complementary products
Le Chat is useful for quick, interactive assistance, while AI Studio is better suited for structured prompt development and repeated experimentation.
Browser-based workflow
Users can work directly in the browser without setting up local software, making it easy to test ideas and integrate AI into everyday tasks.
Academic productivity support
Mistral can support drafting, summarization, exploration of concepts, prompt refinement, and prototyping for research workflows.
Common Use Cases
From images to structured data
From audio to text
Large-scale text analysis
Automated document screening
Data extraction from complex documents
Workflow automation
Code assistance and execution
Model adaptation for specialised research tasks
Multimodal research analysis
Content and material creation
Research design and text refinement
RAG and semantic search
Important Information
Level 1 data only
Currently, only data classified as Level 1 under AAU's data classification policy may be used with Mistral. Data classified as Level 2 or Level 3 must not be entered into Le Chat or AI Studio at this time. We are in the process of conducting risk assessments for handling higher data categories. See the full Terms and Conditions for details.
Choose the right product for the task
Use Le Chat when you need quick conversational assistance. Use AI Studio when you want to test prompts more systematically or work with repeatable AI configurations.
AI output always requires human review
Mistral can save time, but the user remains responsible for checking correctness, bias, citations, compliance, and academic suitability before using generated content.