Claude adds inline interactive charts and diagrams to conversations
Claude has introduced a beta feature that enables the AI to create interactive charts, diagrams, and visualizations directly within chat conversations. This builds on last fall's preview of "Imagine with Claude" and represents a shift from permanent artifacts to temporary, conversation-driven visuals.
How it works
The feature is on by default. Claude will automatically decide when to build a visual, or users can request one directly with prompts like "draw this as a diagram" or "visualize how this might change over time." Once created, users can ask for adjustments or deeper exploration.
Key characteristics
- Inline display: Visuals appear directly in conversation responses rather than in side panels
- Temporary nature: Visuals change or disappear as conversations evolve
- Interactive elements: Users can click around for more details in some visualizations
- Conversation-driven: Designed to aid understanding during discussion rather than create polished artifacts
Examples from the source
- Ask Claude how compound interest works, and it'll provide a curve to interact with
- Ask about the periodic table, and it'll build an interactive visualization where you can click for details
Context and availability
This is part of broader improvements to Claude's response formats. Earlier this year, Claude began using purpose-designed formats for specific topics (recipes now appear with ingredients and steps, weather queries include visuals). The feature is available on all plan types.
📖 Read the full source: HN AI Agents
👀 See Also
Claude Code developer acknowledges adaptive thinking flaw, provides workaround
Boris Charny, creator of Claude Code, confirmed a flaw in the adaptive thinking feature that causes performance degradation. Users experiencing issues even with effort=high settings can use CLAUDE_CODE_DISABLE_ADAPTIVE_THINKING=1 as an interim workaround.
PrismML's Bonsai 1-bit Qwen models tested: 107 t/s generation on 8GB VRAM
Bonsai models from PrismML are 1-bit quantized versions of Qwen3 8B, 4B, and 1.7B that achieve 107 tokens/second generation and >1114 t/s prompt processing on an RTX 4060 with 8GB VRAM, with significantly reduced memory requirements.
Anthropic Removes Gmail Message Body Access from Claude Connector
Anthropic has removed the gmail_read_message and gmail_search_messages tools from the Gmail connector, replacing them with get_thread and search_threads that no longer return message bodies or attachment content.
Tokenmaxxing Is the New Stopwatch: Why Your AI Policy Needs to Be Coherent
Brian Meeker argues against vanity metrics like tokenmaxxing and shares his team's four-point AI policy: no mandate, understand generated code, survive without AI tools, care about teammates and customers.