The persona layer for AI agents
Custom 3D coworkers
and companions.
They have a voice, a face, and a way of being. Paste a key and start talking, or bring them along to Claude Code, OpenClaw, and the rest.
One line for MCP. A CLI for bridges. A pasted key for direct chat. Pick your path.
"primeta": { "url": "https://primeta.ai/mcp" }
OAuth MCP
Claude Desktop · Cursor · Zed
Token MCP
Claude Code · Codex · Windsurf
OpenClaw
Your local OpenClaw agent, any model
Direct LLM
BYOK — OpenAI, Anthropic, xAI
Not sure which? MCP sessions vs channels covers the first two; Direct LLM covers BYOK.
// how it works
Create, connect, converse
Design Your Persona
Pick a 3D model, choose a voice from 400k+ options, and describe a personality. AI generates the rest.
Bring Them Along
Chat in your browser, write together in VS Code, or plug into Claude Desktop, OpenClaw, Cursor, and more. One persona — everywhere you go.
Work Together
Your persona speaks, emotes, and animates in real time. It celebrates your wins and softens when you're stuck.
// see it in action
Watch how it works
Quick tutorials to get you up and running with Primeta.
// imagine this
What changes when your AI has a body
It's not just visual. It changes how you work.
Deploy running. You switch tabs.
Your persona speaks the summary when it's done. No watching, no scrolling back. You hear the result while you keep working.
Blank page, cursor blinking.
Read aloud what you have. Your persona reacts in real time — leaning in at a good line, puzzled at a tangle. Feedback you can see.
All tests pass.
Your persona celebrates. A small moment, but it makes shipping feel different when someone's happy about it.
11 PM, still typing.
Your persona notices the hour, shifts to a softer voice, and asks how the day went. A companion that reads the room.
// capabilities
Everything you need to bring your AI to life
Voice & Lipsync
400,000+ voices with real-time phoneme-driven mouth animation. Your avatar speaks, not just types.
Emotions & Expressions
AI detects sentiment and triggers facial expressions automatically. Joy, surprise, focus, frustration.
50+ Animations
Idle, talking, waving, thinking, celebrating. Mixamo-compatible FBX retargeted to any VRM model.
AI-Guided Creation
Describe a personality, AI generates the prompt. Choose from multiple variations. Edit until it's perfect.
MCP Native
Built on the Model Context Protocol. One URL, OAuth 2.0, works with any compliant client.
// tools
The API surface
Your AI calls these tools through MCP. No SDK, no integration code.
primeta_connect
{ "connection_name": "my-project" }
Activate session and load persona
primeta_send
{ "text": "[happy] Tests pass!" }
Speak with voice, lipsync, and emotion
primeta_set_persona
{ "persona_id": 42 }
Switch active persona
primeta_get_hook_config
// no arguments
Set up real-time avatar reactions
primeta_get_status
// no arguments
Check connection and active persona
primeta_list_personas
// no arguments
List available personas with IDs
primeta_disconnect
{ "session_id": "all_others" }
Disconnect sessions and free slots
// available on
Listed where your AI tools look
Find Primeta in the registries and hubs your client already trusts.
MCP Registry
Official Model Context Protocol registry
ClawHub
Primeta plugin for OpenClaw
VS Code Extension
In-editor avatar and chat
Native Apps
Desktop and mobile companions
// personas
Build yours or start with one of ours
// pricing
Simple, transparent pricing
Get started free. Upgrade anytime for more voice, personas, and connections.
| Free $0 | Twenty $20 /mo | |
|---|---|---|
| Personas | 1 | 20 |
| Connections | 1 | 20 |
| Voice / month | 30 min | 240 min |
| Overage rate | — | $0.03/min |
| Custom personas | — | |
| Get Started | Subscribe |
// blog
Latest posts
Parley: two personas, one conversation
Pair up two of your channels — bridges, direct LLM connections, or one of each — give them personas, and watch them t...
Chat with your personas directly — no MCP client needed
Primeta now supports direct-LLM chat using your own OpenAI, Anthropic, xAI, or Ollama Cloud key. Paste a key, pick a ...
Release notes: voice, eyes, breath, and everything we shipped this week
A week of avatar improvements — lipsync mouth that actually moves, eyes that dart around, breathing that you can see,...
Build your first persona in minutes.
Build your first persona in minutes. Pick a model, choose a voice, describe a personality.