three.ws launches tomorrow (May 28th, 12:01 AM PDT) as the 3D layer for the internet: a fully embodied AI agent (not a flat chatbot) with a real 3D body, 2500+ real-time animations, voice, memory, emotions, DOM-awareness (yes, it literally covers its eyes when you type a password), on-chain identity via Solana Metaplex, and autonomous earning through x402 pay-per-call. One <script> + one <agent-3d> tag and it works on any site Notion, Framer, WordPress, Ghost, or your custom stack. Built open-source on three.js + MCP + Solana, with a browser-native runtime so no backend uploads or installs needed.
Use cases Where would you actually embed a 3D embodied agent first? E-commerce product advisors with real-time 3D demos? Personal sites with a digital twin greeter? Internal team dashboards? Education/tutoring? Something we haven t even thought of?
DOM awareness & privacy The password-eye-covering trick is our favorite demo. What other smart DOM reactions would you want (e.g., reacting to scroll position, form focus, dark mode, or page content)? How far should the privacy layer go?
Monetization & x402 Autonomous earning via pay-per-call feels like the missing piece for agent economies. Have you seen (or built) anything similar? Would you let an agent on your site earn USDC directly, or do you prefer subscription/tip models?
On-chain identity Every agent gets a verifiable Solana Metaplex Core asset (~$0.01 to mint). How do you see portable on-chain agent identities playing out reputation scores, cross-site portability, agent-to-agent payments, or even secondary markets?
Technical depth For devs: Performance on mobile/low-end devices? How does the animation system feel with 2500+ clips? Any gotchas when mixing with React, Next.js, or Svelte? (We ship a vanilla web component, but we want your real-world war stories.)
Comparison & differentiation How does this stack up against other AI agent tools you re using today (Cursor, Lovable, Vapi, etc.)? What s the biggest finally! moment for you with true 3D embodiment vs 2D/chat-only agents?
Roadmap vibes Selfie-to-rigged avatar is already in progress. What other features would make you immediately ship this to production (multiplayer scenes, better voice cloning, more LLM backends, etc.)?
Drop your questions, feature requests, concerns, or even just this would be insane for X below. We re reading every reply and will jump in with answers, demos, and maybe some early access codes for the most interesting threads.Excited to finally share this with the PH community the web is about to feel a lot more alive. Let s build the 3D internet together. three.ws team
<iframe src="https://three.ws/widget#widget=wdgt_demo_turntab&kiosk=true" width="600" height="600" style="border:0;border-radius:12px" allow="autoplay; xr-spatial-tracking" loading="lazy"></iframe>
worth a shot!