Anthropic's Claude has long excelled at coding and text-based tasks, but the company has now joined competitors in adding visual capabilities. On Thursday, Anthropic updated Claude to generate interactive charts, diagrams, and visualizations on demand.
Claude can now create custom charts, diagrams and other visualizations in-line in its responses and then tweak and modify its creations as the conversation develops. The capability operates differently from basic image generation. When producing visual aids, Claude uses HTML code and XML vector graphics, an approach Anthropic likens to giving Claude access to its own whiteboard.
The feature relies on on-the-fly generation of JavaScript code using visualization library Chart.js, HTML, CSS, and related technologies. These are on-demand mini-apps intended to be temporary rather than persistent tools like Claude Artifacts.
What users can do with the feature
Anthropic's canonical example involves asking Claude to explain how compound interest works, with Claude creating an interactive chart that resembles what you would find searching for a compound interest calculator on Google, minus the ads and cookie banners. Users can interact with visuals by clicking buttons, adjusting sliders, or expanding to full screen, then ask Claude to refine or add more detail.
Where Claude was previously limited to text, it can now show step-by-step instructions for tasks such as folding a Nakamura lock plane. Claude can generate flowcharts, structural diagrams showing weight distribution, visual guides for procedural tasks, and data charts for business use, allowing entrepreneurs to chart revenue trends without using a spreadsheet tool.
Availability and limitations
The feature is rolling out as a beta to all Claude users on all plan types. Anthropic warns the software is beta, so users should expect quirks, and the feature is not yet available on mobile. Creating visuals can take considerable time, and when it often takes half a minute to generate a graphic in Claude, a basic web search may feel like a better option.
This feature allows Claude to create new graphical designs on the fly, whereas previous visual responses used pre-set designs. Inline visuals live inside the conversation flow rather than occupying a separate panel, differing from the original Claude Artifacts feature introduced in 2024.
The competitive race for visual AI
OpenAI launched interactive visual tools inside ChatGPT on March 10, covering more than 70 math and science concepts with adjustable sliders. OpenAI's feature focuses mostly on explaining math and science topics to students. A three-day gap between the two launches suggests both companies view interactive visuals as a competitive necessity rather than a differentiator.
Google debuted interactive charts and simulations for Gemini Ultra subscribers at a premium price last December, whereas Anthropic is making its interactive charts available to all users. The contrast reflects different business strategies: Anthropic aims to build its user base on all pricing tiers, while Google reserves advanced visual features for premium subscribers.
Claude has surged to number one among free apps on the U.S. Apple App Store following a recent Pentagon dispute involving OpenAI, giving Anthropic a larger audience for new features. This timing advantage means Anthropic's feature reaches a substantially expanded user base as a flagship capability rather than an incremental update.
These visuals are built using HTML code and SVG vector graphics rather than traditional image generation, which means they load faster and remain interactive across different screen sizes. That technical choice underscores how the AI labs are approaching visualisation not as a cosmetic feature but as a functional tool for user interaction.
The feature remains in beta, and early feedback will likely shape how Anthropic extends visual rendering to mobile devices and expands supported content types in future releases.