First Conversation with Vibe

Updated November 2, 2026 — Le Chat Is Now Vibe. As of May 28, 2026, Le Chat is now Vibe (still at chat.mistral.ai): a single agent for work and code, with your conversations, settings, and subscriptions carried over. The legacy Le Chat experience remains accessible in Chat mode; in Work mode, agents are now skills (invoked with @), memories are the knowledge base, and deep search is the Deep Research skill. Plans are now called Mistral Free, Pro, Team, and Enterprise. This lesson uses the name Vibe; some menus described here may have changed names or locations.

Your First Conversation with Mistral

You’ve created your account and you know the available plans: now it’s time to get hands-on. This lesson walks you through your very first conversation with a Mistral model, from opening the interface to the art of asking a question that gets a good answer. Take the time to follow along as you read—the concepts from previous lessons will click when you actually interact with the model.

Access Vibe

Open your browser and go to chat.mistral.ai. If you’re already logged into the Mistral Console, you’ll be authenticated automatically; otherwise, log in with the credentials you created in the previous lesson. You’ll land on an intentionally minimalist interface: an input field in the center of the page, a model selector placed above or near that field, and a sidebar on the left that keeps your conversation history. Nothing else—and that simplicity is precisely what makes the tool accessible from minute one.

Choose a Model

Before you type anything, take a moment to look at the model selector. Vibe usually offers several options, and this choice matters: it determines the balance between speed and power in the response. Mistral Small 4, the newest model (March 2026), is fast and versatile—it’s often the default choice, and it suits the vast majority of daily tasks. Mistral Large 3 is the most powerful model in the lineup: reserve it for complex tasks that require deep reasoning or multimodal analysis. The reasoning mode of Small 4, finally, handles logic, math, or step-by-step analysis—it activates automatically when the question requires it.

For a first interaction, Mistral Small 4 is the ideal choice: it’s fast, high-performing, and free. You’ll have plenty of opportunity to explore the other models as your needs become clearer.

Ask Your First Question

Click in the input field and get started. If you’re short on inspiration, here are three angles that illustrate three different uses. A simple question to test the model’s general knowledge:

What are the main advantages of artificial intelligence for a French small business?

A writing task to see the model produce a ready-to-use deliverable:

Write a professional email to introduce a new product to an existing client. The product is project management software.

Or a technical question to evaluate its teaching ability:

Explain the difference between machine learning and deep learning, with concrete examples.

Type your question, then press Enter or click the send button.

Understand the Response

The response text appears progressively on the screen, word by word: this is called streaming, and you’ll understand why in the lesson on model architecture. Observe the format of what you receive: the response is structured in paragraphs, lists, or headings depending on the nature of your request, and the model uses markdown to format the text (bold, italic, code blocks). Its length adapts to the complexity of the question—a definition fits in a few lines, an analysis in several sections. And if the response takes a direction that doesn’t suit you, you can stop generation mid-stream rather than waiting for the end.

Interact with the Response

Once the response is displayed, the conversation is just getting started. The most valuable habit to build is the follow-up question: the model keeps the entire exchange in memory, which lets you dig deeper without re-explaining everything. After asking about the advantages of AI for a small business, you can follow up with:

Can you detail the third point? And give concrete examples for a small business in the restaurant sector?

If the response doesn’t satisfy you, ask the model to regenerate its answer: the result will be different, because the model introduces an element of variability in its generations. A Copy button also lets you grab the formatted text to paste into a document. Finally, you can rate each response with a thumbs up or down—this feedback helps Mistral improve its models.

Best Practices for Your First Conversation

The quality of the answer depends directly on the quality of the question, and you learn this quickly. Compare “Write an email” with “Write a sales follow-up email for a B2B client in the industrial sector”: the second phrasing gives the model the context it needs to hit the target. Make it a habit to provide context—your situation, your goal, your audience—, specify the expected format (“as a list”, “in 3 paragraphs”, “in a table”), and indicate the desired language and tone (“in professional French”, “in an informal tone”). These few seconds of upfront effort save you multiple back-and-forth corrections.

Practice

To solidify these habits, complete three exercises on Vibe now:

1. Ask Mistral to explain its own functionality in simple terms
2. Ask it to write a summary of your typical day in 5 points, describing your role
3. Ask it a question in your field of expertise and evaluate the relevance of the response

The third exercise is the most insightful: when you question the model on a topic you know well, you gauge its capabilities—and its limitations—most accurately.

Key Takeaways

• Vibe is accessible on chat.mistral.ai with your Mistral account
• The model selector lets you choose between speed and power
• Mistral Small 4 is the best choice for starting: fast, versatile, free
• The more precise and contextualized your question, the better the response
• You can iterate within a conversation: the model keeps the context of the exchange