Open-source platform for 3D AI agents. Turn text or a photo into a rigged, animated GLB avatar, give it an LLM brain, memory and a wallet, and embed it anywhere with one web component. Remote MCP server, x402 payments, Three.js, browser-native.
JavaScript 3d-generationai-agentsanimationavatardigital-humangenerative-aiglbgltfimage-to-3dllmmcpmcp-servermodel-context-protocolsolanatext-to-3dthreejsweb-componentswebglwebxrx402
<!doctype html>
<html lang="en">
<head>
<script>/* three.ws theme boot — no-flash; runtime in /theme-switcher.js */(function(){try{var m=localStorage.getItem('twx_theme');var l=m==='auto'?(window.matchMedia&&window.matchMedia('(prefers-color-scheme: light)').matches):m==='light';document.documentElement.setAttribute('data-theme',l?'light':'dark');}catch(e){document.documentElement.setAttribute('data-theme','dark');}})();</script>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" />
<title data-i18n="what_is.meta_title">What is three.ws? — One platform for 3D AI agents · three.ws</title>
<meta
name="description"
content="three.ws is one platform for building a 3D AI agent — a character with a face, a brain, and a voice — that you can embed anywhere. See four real ways people use it: support avatars, streamer characters, community mascots, and personal AI agents."
data-i18n-attr="content:what_is.meta_description"/>
<meta name="robots" content="index, follow" />
<link rel="canonical" href="https://three.ws/what-is" />
<meta property="og:type" content="website" />
<meta property="og:site_name" content="three.ws" />
<meta property="og:title" content="What is three.ws? — One platform for 3D AI agents" data-i18n-attr="content:what_is.meta_og_title"/>
<meta
property="og:description"
content="Build a 3D AI agent with a face, a brain, and a voice — and embed it anywhere. Four real use-cases, one start-free flow."
data-i18n-attr="content:what_is.meta_og_description"/>
<meta property="og:url" content="https://three.ws/what-is" />
<meta property="og:image" content="https://three.ws/api/page-og?s=main&t=What+is+three.ws%3F&d=Plain-English+introduction+to+three.ws%3A+what+a+3D+AI+agent+is%2C+real+use-cases%2C+and+where+to+begin.&p=%2Fwhat-is" />
<meta property="og:image:width" content="1200" />
<meta property="og:image:height" content="630" />
<meta name="twitter:card" content="summary_large_image" />
<meta name="twitter:title" content="What is three.ws? — One platform for 3D AI agents" data-i18n-attr="content:what_is.meta_twitter_title"/>
<meta
name="twitter:description"
content="Build a 3D AI agent with a face, a brain, and a voice — and embed it anywhere."
data-i18n-attr="content:what_is.meta_twitter_description"/>
<meta name="twitter:image" content="https://three.ws/api/page-og?s=main&t=What+is+three.ws%3F&d=Plain-English+introduction+to+three.ws%3A+what+a+3D+AI+agent+is%2C+real+use-cases%2C+and+where+to+begin.&p=%2Fwhat-is" />
<link rel="icon" href="/favicon.ico" sizes="any" />
<link rel="icon" type="image/svg+xml" href="/favicon.svg" />
<link rel="icon" type="image/png" sizes="32x32" href="/favicon-32x32.png" />
<link rel="icon" type="image/png" sizes="16x16" href="/favicon-16x16.png" />
<link rel="apple-touch-icon" href="/apple-touch-icon.png" />
<link rel="manifest" href="/manifest.webmanifest" />
<link rel="stylesheet" href="/style.css" />
<link rel="stylesheet" href="/nav.css" />
<link rel="stylesheet" href="/features-landing.css" />
<link rel="stylesheet" href="/what-is.css" />
<link rel="stylesheet" href="/fonts/fonts.css" />
<script
type="module"
src="https://cdn.jsdelivr.net/npm/@google/[email protected]/dist/model-viewer.min.js"
integrity="sha384-wjsPzEO+lXOf7d9csIDfAQpPt2OKuUU3m0K9qsdKIuzX4N8czSF6hdh23eAbZa3p"
crossorigin="anonymous"
></script>
<script type="application/ld+json">
{
"@context": "https://schema.org",
"@graph": [
{
"@type": "WebPage",
"name": "What is three.ws?",
"url": "https://three.ws/what-is",
"description": "A plain-language explainer of three.ws — one platform for building and embedding 3D AI agents — with four concrete use-cases."
},
{
"@type": "FAQPage",
"mainEntity": [
{
"@type": "Question",
"name": "Do I need to know anything about crypto?",
"acceptedAnswer": {
"@type": "Answer",
"text": "No. Creating an avatar, giving it an AI brain, embedding it, and chatting with it are all free and crypto-free. Owning your agent on-chain or charging per chat are optional layers you can turn on later if you ever want them."
}
},
{
"@type": "Question",
"name": "Is it free to start?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Yes. You can build a 3D agent from a selfie or a text prompt, give it a brain, and embed it without paying anything. You only pay if you opt into premium generation or run paid services."
}
},
{
"@type": "Question",
"name": "Do I need 3D modelling or coding skills?",
"acceptedAnswer": {
"@type": "Answer",
"text": "No. The avatar comes from a photo, a sculpt tool, or a text description, with no Blender and no game engine. Embedding is a copy-paste snippet. Developers who want deeper control get an SDK and API, but neither is required."
}
},
{
"@type": "Question",
"name": "Where can I embed or use my agent?",
"acceptedAnswer": {
"@type": "Answer",
"text": "Any website, via the embed widget; on a live stream, via the OBS overlay; in a 3D world on three.ws; or anywhere that accepts a standard glTF 2.0 GLB, since that is the format your avatar exports to."
}
}
]
}
]
}
</script>
<meta property="og:locale" content="en_US">
<meta property="og:image:alt" content="What is three.ws? — three.ws">
<meta name="twitter:site" content="@trythreews">
<script src="/model-viewer-meshopt.js"></script>
</head>
<body class="fl-page">
<header><div id="nav-container"></div></header>
<script src="/nav.js"></script>
<main class="fl-main" id="main-content">
<nav class="fl-breadcrumb" aria-label="Breadcrumb">
<a href="/" data-i18n="what_is.three_ws">three.ws</a>
<span class="fl-breadcrumb-sep" aria-hidden="true">/</span>
<span class="fl-breadcrumb-current" data-i18n="what_is.what_is_three_ws">What is three.ws?</span>
</nav>
<!-- Hero -->
<section class="fl-hero" aria-labelledby="wi-hero-h1">
<div class="fl-hero-inner">
<div class="fl-hero-copy">
<span class="fl-eyebrow"
data-i18n-html="what_is.new_here_start_here"><span class="fl-eyebrow-dot" aria-hidden="true"></span>New here? Start
here</span
>
<h1 id="wi-hero-h1" data-i18n-html="what_is.a_3d_ai_agent_you_can_use">A 3D AI agent<br />you can use anywhere.</h1>
<p class="fl-hero-sub" data-i18n="what_is.three_ws_is_one_platform_for_making">
three.ws is one platform for making a 3D character that has a face, a
brain, and a voice — and putting it to work on your website, your
stream, your community, or your desktop.
</p>
<div class="fl-hero-actions">
<a href="/create" class="fl-btn-primary" data-i18n="what_is.start_free">Start free →</a>
<a href="#use-cases" class="fl-btn-ghost" data-i18n="what_is.see_what_you_can_build">See what you can build</a>
</div>
</div>
<!-- Three layers, one box. The still is a real render of the same GLB
and paints immediately, so the pane is never an empty rectangle:
it is the loading state, and it is also the fallback when the
model-viewer CDN or the GLB never arrives. The label below states
which of the three the visitor is actually looking at. -->
<div class="fl-hero-preview" data-state="loading">
<img
class="fl-preview-still"
src="/media/what-is/hero-avatar.webp"
alt="Sample rigged 3D avatar, an example of the characters you can create on three.ws"
width="339"
height="720"
decoding="async"
fetchpriority="high"
/>
<model-viewer
src="/animations/soldier.glb"
alt="Sample rigged 3D avatar playing its idle animation. Drag to orbit the camera."
autoplay
animation-name="Idle"
auto-rotate
auto-rotate-delay="1500"
rotation-per-second="12deg"
camera-controls
camera-orbit="180deg 82deg auto"
interaction-prompt="none"
shadow-intensity="0.5"
environment-image="neutral"
></model-viewer>
<span class="fl-preview-label">
<span class="fl-preview-state" data-when="loading" data-i18n="what_is.loading_the_sample_avatar">Loading the sample avatar</span>
<span class="fl-preview-state" data-when="ready" data-i18n="what_is.sample_avatar_drag_to_orbit">Sample avatar · drag to orbit</span>
<span class="fl-preview-state" data-when="still" data-i18n="what_is.sample_avatar_still_image_3d_preview_unava">Sample avatar · still image, 3D preview unavailable</span>
</span>
</div>
</div>
</section>
<!-- Plain definition lede -->
<section class="wi-lede fl-section" aria-labelledby="wi-lede-h2">
<h2 id="wi-lede-h2" class="fl-section-label" data-i18n="what_is.in_one_sentence">In one sentence</h2>
<div class="wi-lede-inner">
<p data-i18n-html="what_is.three_ws_lets_anyone_turn_a_selfie">
three.ws lets <strong>anyone</strong> turn a selfie or a description into a
<strong>3D AI agent</strong> — a character you can talk to, animate, and
embed on any website with two lines of code.
<span class="wi-dim"
>No 3D modelling, no game engine, and no crypto required to start.</span
>
</p>
</div>
</section>
<!-- One platform, many pieces -->
<section class="wi-pieces fl-section" aria-labelledby="wi-pieces-h2">
<h2 id="wi-pieces-h2" data-i18n="what_is.one_platform_not_seven_tools">One platform, not seven tools</h2>
<p class="wi-cases-intro" data-i18n="what_is.the_avatar_creator_the_animation_the_ai">
The avatar creator, the animation, the AI chat, the embed widget, and the worlds
are all building blocks for the same agent. Make it once — use it everywhere.
</p>
<div class="wi-pieces-grid">
<a class="wi-piece" href="/create" data-i18n-html="what_is.avatar_creatora_selfie_a_sculpt_or_a">
<span class="wi-piece-ico" aria-hidden="true">◎</span>
<span
><span class="wi-piece-t">Avatar creator</span
><span class="wi-piece-d"
>A selfie, a sculpt, or a text prompt → a rigged 3D character</span
></span
>
</a>
<a class="wi-piece" href="/brain" data-i18n-html="what_is.ai_brainpick_the_model_claude_gpt_and">
<span class="wi-piece-ico" aria-hidden="true">◍</span>
<span
><span class="wi-piece-t">AI brain</span
><span class="wi-piece-d"
>Pick the model — Claude, GPT, and more — and give it
knowledge</span
></span
>
</a>
<a class="wi-piece" href="/overlay-control" data-i18n-html="what_is.animation_motionlip_sync_emotes_mocap_and_an">
<span class="wi-piece-ico" aria-hidden="true">◐</span>
<span
><span class="wi-piece-t">Animation & motion</span
><span class="wi-piece-d"
>Lip-sync, emotes, mocap, and an OBS stream overlay</span
></span
>
</a>
<a class="wi-piece" href="/studio" data-i18n-html="what_is.embeddable_widgettwo_lines_of_html_drop_your">
<span class="wi-piece-ico" aria-hidden="true">▦</span>
<span
><span class="wi-piece-t">Embeddable widget</span
><span class="wi-piece-d"
>Two lines of HTML drop your agent onto any site</span
></span
>
</a>
<a class="wi-piece" href="/play" data-i18n-html="what_is.3d_worldsa_space_your_agent_and_your">
<span class="wi-piece-ico" aria-hidden="true">◇</span>
<span
><span class="wi-piece-t">3D worlds</span
><span class="wi-piece-d"
>A space your agent — and your community — can hang out in</span
></span
>
</a>
<a class="wi-piece" href="/pay" data-i18n-html="what_is.optional_paymentscharge_per_chat_or_give_you">
<span class="wi-piece-ico" aria-hidden="true">◈</span>
<span
><span class="wi-piece-t">Optional payments</span
><span class="wi-piece-d"
>Charge per chat, or give your agent its own on-chain identity</span
></span
>
</a>
</div>
</section>
<!-- Use cases -->
<section class="wi-cases fl-section" id="use-cases" aria-labelledby="wi-cases-h2">
<h2 id="wi-cases-h2" data-i18n="what_is.four_ways_people_use_it">Four ways people use it</h2>
<p class="wi-cases-intro" data-i18n="what_is.same_platform_very_different_jobs_find_the">
Same platform, very different jobs. Find the one that sounds like you — every
step below links to a real, working flow.
</p>
<div class="wi-case-grid">
<article class="wi-case">
<span class="wi-case-who" data-i18n="what_is.for_businesses">For businesses</span>
<h3 class="wi-case-title" data-i18n="what_is.a_support_avatar_on_your_website">A support avatar on your website</h3>
<p class="wi-case-what" data-i18n="what_is.a_branded_3d_greeter_that_answers_your">
A branded 3D greeter that answers your customers' questions 24/7 — a
friendlier face than a chat bubble, embedded in minutes.
</p>
<ol class="wi-case-steps">
<li data-i18n="what_is.create_an_avatar_and_feed_it_your">Create an avatar and feed it your FAQ as its brain.</li>
<li data-i18n="what_is.copy_the_two_line_embed_snippet_from">Copy the two-line embed snippet from Widget Studio.</li>
<li data-i18n="what_is.paste_it_on_your_site_visitors_chat">Paste it on your site — visitors chat with it instantly.</li>
</ol>
<div class="wi-case-actions">
<a href="/create" class="fl-btn-primary" data-i18n="what_is.build_a_support_avatar">Build a support avatar →</a>
<a href="/studio" class="fl-btn-ghost" data-i18n="what_is.see_widget_studio">See Widget Studio</a>
</div>
</article>
<article class="wi-case">
<span class="wi-case-who" data-i18n="what_is.for_streamers_creators">For streamers & creators</span>
<h3 class="wi-case-title" data-i18n="what_is.an_on_stream_character_that_s_you">An on-stream character that's you</h3>
<p class="wi-case-what" data-i18n="what_is.a_3d_avatar_for_youtube_or_twitch">
A 3D avatar for YouTube or Twitch that lip-syncs your mic and emotes on
hotkeys — go on camera as a character without a pro VTuber rig.
</p>
<ol class="wi-case-steps">
<li data-i18n="what_is.scan_your_selfie_into_a_rigged_3d">Scan your selfie into a rigged 3D avatar.</li>
<li data-i18n="what_is.add_it_to_obs_as_a_browser">Add it to OBS as a browser overlay.</li>
<li data-i18n="what_is.it_lip_syncs_your_voice_and_reacts">It lip-syncs your voice and reacts to hotkeys, live.</li>
</ol>
<div class="wi-case-actions">
<a href="/create/selfie" class="fl-btn-primary" data-i18n="what_is.scan_your_avatar">Scan your avatar →</a>
<a href="/overlay-control" class="fl-btn-ghost" data-i18n="what_is.open_the_obs_overlay">Open the OBS overlay</a>
</div>
</article>
<article class="wi-case">
<span class="wi-case-who" data-i18n="what_is.for_communities">For communities</span>
<h3 class="wi-case-title" data-i18n="what_is.a_mascot_and_a_world_to_gather">A mascot and a world to gather in</h3>
<p class="wi-case-what" data-i18n="what_is.a_3d_character_that_represents_your_communit">
A 3D character that represents your community — and its own little world
people can drop into and hang out, with an optional token if you want
one.
</p>
<ol class="wi-case-steps">
<li data-i18n="what_is.forge_or_create_your_mascot_character">Forge or create your mascot character.</li>
<li data-i18n="what_is.drop_it_into_its_own_3d_world">Drop it into its own 3D world people can explore.</li>
<li data-i18n="what_is.optionally_give_it_a_token_and_on">Optionally give it a token and on-chain identity.</li>
</ol>
<div class="wi-case-actions">
<a href="/forge" class="fl-btn-primary" data-i18n="what_is.create_your_mascot">Create your mascot →</a>
<a href="/play" class="fl-btn-ghost" data-i18n="what_is.explore_worlds">Explore Worlds</a>
</div>
</article>
<article class="wi-case">
<span class="wi-case-who" data-i18n="what_is.for_builders_tinkerers">For builders & tinkerers</span>
<h3 class="wi-case-title" data-i18n="what_is.a_personal_ai_agent_with_a_face">A personal AI agent with a face</h3>
<p class="wi-case-what" data-i18n="what_is.your_own_assistant_you_can_actually_look">
Your own assistant you can actually look at and talk to — pick its
brain, give it a personality, then chat with it or call it from the API.
</p>
<ol class="wi-case-steps">
<li data-i18n="what_is.create_an_agent_and_choose_its_brain">Create an agent and choose its brain — Claude, GPT, and more.</li>
<li data-i18n="what_is.give_it_a_personality_and_the_knowledge">Give it a personality and the knowledge it needs.</li>
<li data-i18n="what_is.chat_with_it_embed_it_or_drive">Chat with it, embed it, or drive it from the API.</li>
</ol>
<div class="wi-case-actions">
<a href="/create" class="fl-btn-primary" data-i18n="what_is.make_your_agent">Make your agent →</a>
<a href="/chat" class="fl-btn-ghost" data-i18n="what_is.try_the_chat">Try the chat</a>
</div>
</article>
</div>
</section>
<!-- FAQ -->
<section class="fl-faq fl-section" aria-labelledby="wi-faq-h2">
<h2 id="wi-faq-h2" data-i18n="what_is.common_questions">Common questions</h2>
<div class="fl-faq-list">
<div class="fl-faq-item">
<h3 class="fl-faq-h">
<button class="fl-faq-q" type="button" id="wi-faq-q1" aria-expanded="false" aria-controls="wi-faq-1">
<span data-i18n="what_is.do_i_need_to_know_anything_about_crypto">Do I need to know anything about crypto?</span>
<svg
class="fl-faq-chevron"
viewBox="0 0 16 16"
fill="none"
aria-hidden="true"
>
<path
d="M4 6l4 4 4-4"
stroke="currentColor"
stroke-width="1.5"
stroke-linecap="round"
/>
</svg>
</button>
</h3>
<div class="fl-faq-a" id="wi-faq-1" role="region" aria-labelledby="wi-faq-q1">
<p data-i18n="what_is.no_creating_an_avatar_giving_it_an">
No. Creating an avatar, giving it an AI brain, embedding it, and
chatting with it are all free and crypto-free. Owning your agent
on-chain or charging per chat are optional layers you can turn on
later if you ever want them.
</p>
</div>
</div>
<div class="fl-faq-item">
<h3 class="fl-faq-h">
<button class="fl-faq-q" type="button" id="wi-faq-q2" aria-expanded="false" aria-controls="wi-faq-2">
<span data-i18n="what_is.is_it_free_to_start">Is it free to start?</span>
<svg
class="fl-faq-chevron"
viewBox="0 0 16 16"
fill="none"
aria-hidden="true"
>
<path
d="M4 6l4 4 4-4"
stroke="currentColor"
stroke-width="1.5"
stroke-linecap="round"
/>
</svg>
</button>
</h3>
<div class="fl-faq-a" id="wi-faq-2" role="region" aria-labelledby="wi-faq-q2">
<p data-i18n="what_is.yes_you_can_build_a_3d_agent">
Yes. You can build a 3D agent from a selfie or a text prompt, give it
a brain, and embed it without paying anything. You only pay if you
opt into premium generation or run paid services.
</p>
</div>
</div>
<div class="fl-faq-item">
<h3 class="fl-faq-h">
<button class="fl-faq-q" type="button" id="wi-faq-q3" aria-expanded="false" aria-controls="wi-faq-3">
<span data-i18n="what_is.do_i_need_3d_modelling_or_coding">Do I need 3D modelling or coding skills?</span>
<svg
class="fl-faq-chevron"
viewBox="0 0 16 16"
fill="none"
aria-hidden="true"
>
<path
d="M4 6l4 4 4-4"
stroke="currentColor"
stroke-width="1.5"
stroke-linecap="round"
/>
</svg>
</button>
</h3>
<div class="fl-faq-a" id="wi-faq-3" role="region" aria-labelledby="wi-faq-q3">
<p data-i18n-html="what_is.no_the_avatar_comes_from_a_photo">
No. The avatar comes from a photo, a sculpt tool, or a text
description, with no Blender and no game engine. Embedding is a
copy-paste snippet. Developers who want deeper control get an
<a href="/docs" class="fl-faq-link">SDK and API</a>, but neither is
required.
</p>
</div>
</div>
<div class="fl-faq-item">
<h3 class="fl-faq-h">
<button class="fl-faq-q" type="button" id="wi-faq-q4" aria-expanded="false" aria-controls="wi-faq-4">
<span data-i18n="what_is.where_can_i_embed_or_use_my">Where can I embed or use my agent?</span>
<svg
class="fl-faq-chevron"
viewBox="0 0 16 16"
fill="none"
aria-hidden="true"
>
<path
d="M4 6l4 4 4-4"
stroke="currentColor"
stroke-width="1.5"
stroke-linecap="round"
/>
</svg>
</button>
</h3>
<div class="fl-faq-a" id="wi-faq-4" role="region" aria-labelledby="wi-faq-q4">
<p data-i18n-html="what_is.any_website_via_the_embed_widget_on">
Any website, via the
<a href="/studio" class="fl-faq-link">embed widget</a>; on a live
stream, via the OBS overlay; in a 3D world on three.ws; or anywhere
that accepts a standard glTF 2.0 GLB, since that is the format your
avatar exports to.
</p>
</div>
</div>
</div>
</section>
<!-- Final CTA -->
<section class="fl-cta-band" aria-labelledby="wi-cta-h2">
<div class="fl-cta-inner">
<h2 id="wi-cta-h2" data-i18n="what_is.make_one_in_a_few_minutes">Make one in a few minutes</h2>
<p data-i18n="what_is.bring_a_selfie_or_just_an_idea">
Bring a selfie or just an idea. Building your first agent is free — no
account needed to try.
</p>
<div class="fl-cta-actions">
<a href="/create" class="fl-btn-primary" data-i18n="what_is.start_free_2">Start free →</a>
<a href="/features" class="fl-btn-ghost" data-i18n="what_is.tour_every_feature">Tour every feature</a>
</div>
</div>
</section>
</main>
<script>
document.querySelectorAll('.fl-faq-q').forEach((btn) => {
btn.addEventListener('click', () => {
const expanded = btn.getAttribute('aria-expanded') === 'true';
btn.setAttribute('aria-expanded', String(!expanded));
const panel = document.getElementById(btn.getAttribute('aria-controls'));
if (panel) panel.classList.toggle('open', !expanded);
});
});
// Hero preview state. model-viewer is a third-party module served from a
// CDN, so three things can go wrong independently: the script never
// arrives, the custom element upgrades but the GLB fails, or a device
// with no WebGL never paints. All three land on `still`, which keeps the
// real render on screen and stops the label promising an orbit control
// that is not there. `ready` is only set once the model is actually up.
(function () {
const shell = document.querySelector('.fl-hero-preview');
if (!shell) return;
const viewer = shell.querySelector('model-viewer');
const settle = (state) => {
if (shell.dataset.state === 'ready') return;
shell.dataset.state = state;
};
viewer.addEventListener('load', () => (shell.dataset.state = 'ready'), { once: true });
viewer.addEventListener('error', () => settle('still'), { once: true });
const cdn = document.querySelector('script[src*="model-viewer"]');
if (cdn) cdn.addEventListener('error', () => settle('still'), { once: true });
// Backstop for the failure modes that fire no event at all: a request
// that hangs rather than errors, or a browser that refuses the WebGL
// context. Generous enough not to trip a slow phone on the 660 KB GLB.
setTimeout(() => settle('still'), 12000);
})();
</script>
<div id="footer-container"></div>
<script src="/footer.js"></script>
<!-- Runtime i18n: detects locale, swaps annotated copy, mounts <lang-switcher>. -->
<script type="module" src="/i18n.js"></script>
</body>
</html>