Anthropic has launched a new feature for Claude that embeds interactive visualisations directly into conversations. When the AI determines a visual representation would help clarify a concept, it now inserts charts, diagrams, and other graphics in-line rather than opening them in a separate side panel.
The capability ranges from practical to playful. Claude can generate an interactive periodic table where users click individual elements for additional information. It can also create visuals explaining how weight distributes through a building structure. More broadly, whenever a user asks about a topic where a visual could add value, Claude evaluates whether to generate one automatically, or users can request a specific diagram, chart, or table directly.
The rollout differs from Claude's existing "artifacts" feature, which creates persistent documents, tools, and apps that users can modify and download. The new inline visualisations exist within the flow of conversation itself, changing or disappearing as the conversation progresses, though users can request modifications on the fly.
This announcement arrives amid a broader industry shift toward visual learning tools. OpenAI introduced dynamic visual explanations in ChatGPT earlier this week, allowing users to interact with formulas and variables in real time across more than 70 mathematics and science concepts. Google launched interactive images in Gemini in November, enabling students to tap directly on diagram components to unlock detailed explanations and follow-up questions.
The competitive urgency is clear. Companies are racing to differentiate their AI assistants in education and data analysis, two high-value markets. OpenAI notes that more than 140 million people use ChatGPT weekly for maths and science help. Google emphasises that interactive images transform passive learning into active exploration, addressing the limitations of static textbook diagrams. Anthropic's latest update signals its determination not to be outpaced in these domains.
There is a meaningful distinction in approach. Claude's visualisations are contextual and conversational, generated on demand within the flow of dialogue. OpenAI's focus skews heavily toward educational scaffolding: step-by-step guidance that encourages learners to find answers themselves rather than simply presenting solutions. Google's interactive diagrams emphasise exploratory learning, where clicking unlocks contextual explanations embedded in the visual itself.
For business users and data analysts, the practical implications are significant. The ability to generate charts and diagrams without leaving the conversation reduces friction; users no longer need to toggle between a chat interface and a side panel to interact with visuals. For students, these tools reflect a philosophical shift across the industry: AI is being positioned less as an answer engine and more as a tutoring partner that helps people think through complex ideas.
The technology does carry constraints worth noting. None of these tools yet generate truly complex, publication-quality visualisations at scale. Claude's inline visuals are interactive but contextual; they serve the conversation rather than functioning as standalone documents. OpenAI's feature is limited to pre-defined concept sets. Google's approach requires careful content curation to maintain educational accuracy. The integration of AI-generated visuals into existing workflows and educational systems remains a live challenge for all three companies.
What is unfolding is a commoditisation race around visual explanation. Within weeks, OpenAI, Google, and Anthropic have all launched comparable features. The window in which any single company could claim a first-mover advantage in AI-generated visualisations has effectively closed. The question now shifts: which company can best integrate visual learning into its broader ecosystem, and which can do so most reliably at scale? For users, the rapid convergence suggests that visual AI features will soon become standard across all major AI assistants rather than differentiation. That is probably healthy for the market, though it does mean companies will need to compete on reliability, integration, and pedagogical soundness rather than novelty alone.