01 NVIDIA Inception · Member since July 2026

three.ws

The platform for 3D AI agents — render, embody, own, embed.

Drop a 3D model, give it an LLM brain and a voice, mint its identity on-chain, and embed it anywhere — in the browser. No plugins. No installs.

three.ws · LIVE Apache 2.0 · Open source Browser-native NVIDIA Inception member
Nicholas Resendez · nich@three.wsLive demo: three.ws/create
02 Mission

We build the runtime & identity layer for embodied AI agents.

Today's AI agents are text in a chat box — no face, no body, no persistent identity, no place in the 3D/spatial web. As AI moves from chat to digital humans, there is no open platform to create, run, and distribute them.

three.ws is that platform.

Mission
03 Product · all shipped & live

Four primitives

PrimitiveWhat it does
RenderWebGL 2.0 viewer (three.js r184) loads & validates glTF/GLB with Draco, KTX2, Meshopt — zero server upload✅ Live
EmbodyLLM brain runs a tool-loop: listen → reason → act (animate, express, skills, memory) → speak with real-time lip-sync✅ Live
OwnOn-chain identity — ERC-8004 on 15+ EVM chains, Metaplex Core on Solana, signed action history & reputation✅ Live
Embed<agent-3d> web component + 5 widgets, OpenGraph/oEmbed, versioned CDN — agent into any site✅ Live

Plus a shipped social layer: multiplayer 3D worlds (/play, /club, /city), presence & DMs, voice/pose/mocap studios, and a paid agent-to-agent economy (x402).

Product
04 Proof of concept · see it working

It's live, today

Talking agent

A live agent that listens, reasons, animates, and speaks in real time. three.ws/agents

Embed in 60s

<agent-3d src="…"> — WYSIWYG editor at three.ws/embed.

Selfie → 3D avatar

3 photos → GPU reconstruction → riggable GLB. How it works

Multiplayer worlds

Shared 3D space per token — peer avatars, chat, voxel building. three.ws/play

Reviewers approve faster with a live product + public GitHub + demo link. We lead with all three.

Demo
05 Technology · the agent runtime

How an agent thinks & acts

Event-bus architecture decouples body, brain, voice, memory & skills — every action flows through typed events with a 200-action replay buffer.

Per-turn tool-loop: STT/text in → system prompt assembled from manifest + recalled memory + installed skills → LLM called with full tool set → up to 8 tool iterations → spoken response with ARKit-52 (52-blendshape) lip-sync.

Skills are composable bundles loadable from IPFS/HTTP, with payment-gated trust modes via x402.

Tools per turn

play_clip setExpression lookAt speak remember skill

Published SDKs

@three-ws/avatar · /avatar-schema · /avatar-cli · /solana-agent · @three-ws/mcp-server

Runtime
06 Technology · where the compute lives

A fundamentally GPU-bound product

1 · Inference

Claude primary; open models (Llama 3.3 70B, Mistral, Qwen) via OpenRouter/Groq & Livepeer decentralized GPU. Keys never client-side.

2 · Generative 3D

Selfie/text → textured, rigged avatar. TRELLIS, Hunyuan3D, UniRig — diffusion+transformer, ~30–90s on A100/H100.

3 · Rendering

WebGL 2.0 on every client GPU; WebGPU on the roadmap. PBR, skinned mesh, morph targets, BVH raycasting.

Every avatar created and every open-model inference burns GPU cycles — demand scales linearly with users.

Compute
07 ★ NVIDIA alignment · the make-or-break slide

Built for the NVIDIA stack

three.ws is digital humans + embodied agents — exactly the workload NVIDIA's AI stack targets.

three.ws componentNVIDIA technology
Avatar lip-sync (ARKit-52), facial animationNVIDIA ACE — Audio2Face
Speech-to-text + text-to-speechNVIDIA Riva
LLM agent brains served at scaleNVIDIA NIM + TensorRT-LLM
Selfie/text → 3D (TRELLIS, Hunyuan3D, UniRig)CUDA inference · NIM for 3D / Edify 3D
3D asset pipeline (glTF/GLB, retarget)Omniverse / OpenUSD interop
Decentralized inference network (roadmap)NVIDIA-accelerated node operators

The plan is concrete: Inception GPU credits run avatar-gen + open-model inference, and Inception support carries our digital-human stack onto ACE + NIM.

NVIDIA fit
08 Architecture · production-grade, full-stack

The system

Frontend

Vanilla JS + Vite · three.js r184 WebGL 2.0 · standalone Svelte chat app

Backend

Vercel serverless (Node 24) · Neon Postgres · Cloudflare R2 · Upstash Redis · Colyseus multiplayer (Fly.io)

GPU tier

Replicate · GCP Cloud Run (NVIDIA GPUs) · HF Spaces — pluggable, env-selected generative-3D inference

On-chain & interop

ERC-8004 (15+ EVM chains) · Metaplex Core (Solana) · x402 payments · OAuth 2.1 · MCP over HTTP · OpenAPI 3.1

Architecture
09 Market opportunity

Why now

Who it's for: creators & brands embedding talking 3D agents; AI builders who need a face/voice/body; web3 & consumer-social communities.

The convergence: capable LLMs + fast image/text→3D + on-chain identity — none existed together 18 months ago. AI is moving from chat to embodied digital humans, and the tooling is fragmented and closed.

Our wedge: the only open, browser-native, on-chain agent platform — vs. closed digital-human SaaS and text-only agent frameworks.

[$XX B]
[digital-human / conversational-AI TAM by 20XX — cite a named analyst]

Reviewers expect one credible figure, not a guess.

Market
10 Traction & milestones

Shipped, in production

  • Full agent runtime — viewer, lip-sync, emotion, skills, memory
  • On-chain identity on 15+ EVM chains + Solana
  • <agent-3d> + 5 widgets on versioned CDN
  • Multiplayer worlds, presence, DMs
  • x402 agent economy; MCP tools on x402scan + MCP Registry
  • Selfie→3D GPU avatar pipeline
  • Partner listings: AWS · Alibaba Cloud · BNB Dappbay
[#]
[agents created]
[#]
[MAU / sessions]
[#]
[embeds live]
[$]
[x402 / token volume]

Roadmap: ACE/NIM migration → personalization & reputation markets → open decentralized inference network.

Traction
11 Business model

How it makes money

Generative-3D credits

Selfie/text→avatar generation — GPU cost + margin. Primary variable cost is GPU.

x402 agent economy

Platform fee on agent-to-agent paid skill calls, asset downloads, and royalties.

Embeds / Pro

Premium widgets, custom voices, private skills, brand analytics.

On-chain primitives

Name service (.threews.sol), launchpad, token & community tooling.

With Inception: GPU credits + ACE/NIM efficiency gains directly improve margin and let generation scale with demand.

Model
12 Team

Who's building it

Nicholas Resendez

Founder — product, engineering, and the agent platform. Builds three.ws in public at github.com/nirholas.

Catherine Macoviak

Developer / Engineer — platform and product engineering.

A team of six

Engineering-heavy, shipping daily across real-time 3D, agents, and infrastructure.

Evidence of execution: a production platform spanning real-time 3D, LLM agent runtimes, GPU inference orchestration, smart contracts on 15+ chains, and multiplayer infra — live at three.ws.

Team
13 Membership

What we're doing with Inception

  1. GPU credits — running generative-3D avatar inference + open-model LLM serving at scale
  2. Technical guidance — migrating our digital-human stack to ACE (Audio2Face + Riva) and NIM / TensorRT-LLM
  3. Ecosystem access — Omniverse/OpenUSD interop, co-marketing, the Inception network

three.ws — embodied AI, owned by its creators.

three.ws · github.com/nirholas/three.ws · three.ws/createnich@three.ws