Compare Multiple Images in a Conversation

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

Beyond a Single Image

Analyzing a single image is powerful, but Mistral’s multimodal models truly shine when you send multiple images in the same conversation. The model can then compare, contrast, and draw connections between images—just as you would by placing two documents side by side on your desk. This capability turns Vibe into a decision-support tool: choosing between two options, spotting a difference, or verifying consistency.

How to Send Multiple Images

In Vibe, you can attach multiple images to a single message. You have three options: select multiple files via the upload button while holding Ctrl or Cmd to choose several, drag and drop multiple files simultaneously into the chat area, or send images one after another in consecutive messages—the model retains the conversation context. The first method, sending multiple images in a single message, is generally most effective for direct comparisons because the model sees the entire set at once rather than reconstructing the comparison from memory.

Types of Comparisons

The first and most intuitive use case is before/after: showing two states of the same thing. Think two versions of a design—mockup A versus mockup B—a location at two different times, last month’s chart versus this month’s, or a new product next to the same worn product. The standard prompt is simple: “Compare these two images. What are the main differences?” The model then scrutinizes both visuals and lists what has changed.

The second use case is choosing between options. Hesitating between several products, layouts, or colors? Upload the candidates and ask, “Which of these two covers looks more professional?”, “Compare these three logos; which is the most legible?”, or “Which kitchen layout seems most functional to you?” The model provides a reasoned analysis of each option, often citing criteria you might not have considered yourself. Its real value lies not in deciding for you, but in objectifying the comparison.

The third use case is trend analysis. By uploading a series of charts or chronological screenshots—for example, sales from January to April, or three monthly dashboards—you can ask the model to identify the overall evolution and spot anomalies. What would take a long time to read through manually boils down to a single question.

Finally, consistency checks ensure that multiple documents tell the same story: Do these two invoices come from the same supplier? Do the figures in this table match those in the chart? Are there differences between these two versions of a contract? In a professional context, this is a discreet but highly valuable use case.

Creative Use Cases

Beyond classic comparisons, users find inventive ways to leverage multi-image analysis. In interior design, show a photo of your room and an image of a piece of furniture found online, then ask if the styles match. In fashion, upload a photo of your clothes and request outfit suggestions. In the kitchen, photograph the ingredients available in your fridge and ask for a recipe that uses them. In education, a student can upload an exercise prompt and their handwritten answer to get commented feedback. And when traveling, compare two hotels based on their photos to choose the best one—the model will notice layout details that marketing descriptions overlook.

Best Practices for Multi-Image

A few habits significantly improve results. Number your images in your message—“Image 1 is the current design, Image 2 is the proposal. Compare them.”—to help the model structure its response and avoid confusion about “the first” and “the second.” Limit the number of images: two to four per message works well; beyond that, the model may lose precision on individual details. Prefer consistent resolution, with images of similar size and quality, for a fair comparison. And ask a specific question: “What differences do you see?” is always better than sending two images without commentary.

Limitations of Comparison

Keep the nature of the analysis in mind: the model does not overlay images pixel by pixel; its approach remains semantic—it compares the meaning of what it sees, not exact pixels. Subtle differences in color or size may therefore escape it, and on very similar images—two versions of the same document where only a comma has moved—it may miss the detail. It is not a visual diff tool. Finally, the number of images per message has a technical limit: in practice, stay under 5 images for optimal results.

Hands-On Practice

Here are three exercises to gauge the tool’s capabilities. Take two photos of the same object from different angles and ask the model to describe the perspective differences. Next, take two screenshots of the same website, one on mobile and one on desktop, and request a usability comparison. Finally, upload a chart and the data table that served as its source, and ask the model to verify consistency between the two—this tests the most demanding type of comparison: crossing visual and numerical data.

Key Takeaways

• You can send multiple images in a single message for direct comparisons
• The model excels at before/after comparisons, choosing between options, and trend analysis
• Number your images and ask specific questions for better results
• Stay under 5 images per message and prefer clear, well-framed images