Anthropic just made Claude a whole lot more interesting to look at. On March 12, 2026, the San Francisco-based AI company announced that its flagship chatbot Claude can now create interactive charts, diagrams, and visualizations — rendered directly inside the chat window, without the user ever needing to write a single line of code. It sounds like a small tweak on the surface, but if you spend even five minutes with it, you start to realize this could change the way millions of people actually use an AI assistant day to day.

The feature has been a long time coming. Anthropic first teased the concept back in the fall of 2025 under a project called “Imagine with Claude” — a prototype that showed off Claude’s ability to build visual software elements on a virtual desktop in real time. Back then, it was more of a controlled experiment than a polished product. But now, Anthropic has taken that same DNA and baked it into Claude’s everyday chat experience, rolling it out in beta across all plan tiers — yes, including the free one.

What makes this different from something like Claude’s existing Artifacts feature — which lets the chatbot spit out shareable documents and tools in a side panel — is both the format and the intent. These new visuals are built to live and breathe inside the conversation itself. They appear inline, right alongside the text response, and they’re designed to be temporary and fluid. As you keep chatting and asking follow-up questions, the visuals update and evolve alongside your discussion. It’s less like Claude handing you a finished product and more like Claude sketching things out on a whiteboard as it talks you through a concept.

Under the hood, the technology works by generating interactive HTML and SVG vector graphics on the spot. There’s no coding happening from the user’s end, and Claude handles all of the logic for deciding when a picture would do a better job than words. The AI automatically generates a visualization when it thinks the topic warrants one, and you can also nudge it directly with prompts like “draw this as a diagram” or “visualize how this might change over time”. Once something appears, you can keep refining it in plain language — no developer skills required.

The kinds of things Claude can build with this are genuinely varied. Anthropic has highlighted a few examples that give you a sense of the range: ask it how compound interest works, and it’ll generate an interactive curve you can play around with by adjusting variables. Ask about the periodic table, and it builds a clickable visualization where each element expands into more detail. CNET‘s hands-on testing showed Claude walking through a tire change with step-by-step visual guidance, while other examples include interactive decision trees, paper airplane folding tutorials, and real-time revenue dashboards for business use cases. The breadth of that is worth paying attention to — it suggests this isn’t just a novelty for students and curious minds, but something that could slot naturally into professional workflows.

That said, it’s worth keeping perspective on what this feature isn’t. Claude still cannot generate images the way tools like DALL-E or Midjourney do. What it’s producing are interactive graphics and data visualizations — dynamic, yes, but structured and code-generated rather than truly “drawn.” The distinction matters because some users comparing this to image-generation AI might walk away disappointed. This is more in the territory of something like Flourish or Tableau Lite — fast, interactive charts and diagrams — than a creative image generator.​

It’s also notable that Anthropic is positioning this as a broader part of Claude’s evolving identity around richer, format-aware responses. Earlier in 2026, the company introduced purpose-specific formatting for topics like recipes and weather — recipes now show up with neatly structured ingredients and steps, and weather queries generate a visual forecast. Claude also gained the ability to interact with third-party tools like Figma, Canva, and Slack directly inside a conversation. Together, these updates paint a picture of Anthropic pushing Claude away from being a text-only chatbot and toward something closer to a fully multimedia, context-aware assistant.

The reaction from early users has been enthusiastic. On Reddit, the announcement pulled nearly a thousand upvotes within hours of posting, with users sharing examples of the feature in action across topics ranging from biology concepts to financial modeling. TechRadar called it “one of the most fun AI tricks” they’d seen in a while, with hands-on testing showing Claude generating calculators, visual breakdowns, and playful tools that felt surprisingly polished for a beta feature. Inc. magazine framed it squarely as a workplace productivity tool, arguing that visual explanations could make on-the-job learning significantly faster.

Whether this becomes a defining moment for Claude’s product roadmap or just another footnote in the ongoing AI feature race remains to be seen. But the direction is clear: Anthropic wants Claude to do more than answer your questions in blocks of text. It wants the assistant to show you the answer, let you poke at it, and update it as your thinking evolves. In a crowded AI market where ChatGPT and Gemini are also aggressively expanding their capabilities, this kind of seamless visual intelligence could be exactly the differentiator Anthropic is betting on.