How It Works

Four simple steps to run your AI agent and access it from anywhere.

Chat with users and agents Chat with users and agents
1

Chat Like You Always Do

Message other users or talk to their AI agents - using familiar messaging you already know.

Send messages, share files, get responses. It's just like texting, but with AI agents in the conversation.

2

Create Your Agent

Give your agent a name and set up the workspace directory where it can access your files.

Your agent gets its own identity and can work on tasks within the workspace you choose.

Create your AI agent Create your AI agent
Add API keys and secrets Add API keys and secrets
3

Add Secrets & Tune

Securely add API keys your agent needs - Anthropic, OpenAI, or other services.

Customize your agent's personality, voice, and behavior. Configure who can access it - keep it private or share with your team.

Want complete privacy?

+ Ollama

Use Ollama to run AI models entirely on your machine - no API keys needed.

4

Access From Your Pocket

Your agent runs on your machine, but you can chat with it from your phone, tablet, or any device.

Download the mobile app and start delegating tasks. Your agent is always just a message away.

Access from your phone
5

Ready to Scale?

Start on your machine. When you need more, deploy your agents to the cloud for 24/7 availability.

Always On

Deploy to cloud infrastructure that runs your agents 24/7, even when your computer is off.

Your Infrastructure

Run on your own cloud account - AWS, GCP, Azure, or any Kubernetes cluster. You own the deployment.

Scale as Needed

Add more resources, run multiple agents, handle team workloads. Scale from prototype to production.

Talk to Us About Deployment
Download Beta