Claude now builds interactive charts right inside your chat

Explaining complex concepts through walls of text has always been the weakest part of using AI chatbots. While side-panel features like Artifacts solved part of the problem, they often felt too heavy for quick questions.

Anthropic has now rolled out a beta feature that lets Claude generate interactive charts, diagrams, and other visualizations right inside the conversation stream. The update is enabled by default and available across all Claude plan types.

Scratchpad visuals that evolve with the chat

This update traces back to the Imagine with Claude preview from last fall, which aimed to create real-time visuals without requiring any code.

Unlike Claude's existing Artifacts feature, which creates persistent documents and tools in a separate side panel for sharing or exporting, these new visuals serve an entirely different purpose. They appear inline within the chat bubbles as temporary learning aids that change, adapt, or disappear as the conversation moves forward.

For instance, if you ask Claude how compound interest works, it can generate an interactive curve that you can adjust on the fly. If you ask about the periodic table, it creates a clickable table that reveals element details on demand. In another example reported by The Verge, asking how structural weight travels through a building prompts Claude to illustrate load distribution visually.

Claude decides automatically when a visual will help explain an answer, but users can also prompt it directly with commands like draw this as a diagram or visualize this trend over time. Once an element appears, you can ask Claude to tweak parameters or drill down deeper.

Earlier this year, Claude started rolling out tailored layouts for topics like cooking recipes and weather, alongside deeper integrations with apps like Figma, Canva, and Slack. Meanwhile, rivals are moving in the exact same direction. OpenAI just launched interactive math and science visualizations in ChatGPT earlier this week, and Google Gemini has introduced similar interactive educational imagery.

Why it matters

We think Anthropic made a very sharp distinction here between building a final deliverable and sketching a scratchpad explanation.

Artifacts are fantastic when you want a finished app or a dashboard to keep. But when you are just trying to wrap your head around physics or an amortization schedule, opening a dedicated side dock feels disruptive. By treating visuals as disposable, conversational objects that evolve and fade naturally, Claude makes interactive learning feel much lighter.

For everyday users and students, being able to directly play with parameters instead of reading five paragraphs of technical jargon is an immediate upgrade. There is no barrier to entry either, since it is turned on by default for all tiers.

That said, the success of this approach hinges on how well Claude judges context. If the model gets overly eager and generates bulky interactive blocks for simple queries, it risks cluttering the conversation feed. The race among Anthropic, OpenAI, and Google is no longer just about whose model is smarter, but whose interface can morph into the right interactive tool at the exact right second.