Deploy your own AI agent in 60 seconds — no infrastructure knowledge required.
ZeroOne is a managed platform for creating, deploying, and running personal AI agents. Each agent gets its own identity, memory, and can be connected to Telegram, Discord, or Slack — all configured through a simple dashboard.
- Create a personal AI assistant that knows who you are, how you communicate, and remembers past conversations
- Connect it to your channels — chat with your agent directly on Telegram, Discord, or Slack
- Choose any AI model — Claude, GPT-4, Gemini, Llama, and 15+ others via OpenRouter, Anthropic, OpenAI, Ollama, Groq
- Manage multiple agents — deploy separate agents for different roles, teams, or workflows
- Control everything from a dashboard — start, stop, restart, edit, or delete agents without touching a terminal
- Fill out the form — give your agent a name, pick a model, write its personality
- Connect channels — add a Telegram bot token, Discord bot, or Slack app
- Deploy — ZeroOne spins up a dedicated container for your agent in ~60 seconds
- Start chatting — your agent is live on your chosen channels and the web dashboard
No server setup. No Docker knowledge needed. No API key juggling.
- One-click deploy — agent goes from form to running in ~60 seconds
- Edit identity, memory settings, and channel connections at any time
- Visual restart indicator when config changes are being applied
- Live agent status (Starting, Running, Stopped, Error)
- Start / Stop / Restart / Delete from the dashboard
- Custom agent name, persona, communication style, and timezone
- Persistent memory — your agent remembers conversations across sessions
- User profile — the agent knows your name and preferences
- Telegram — connect a bot token, agent replies in your Telegram DMs or group
- Discord — connect a bot to a server, chat in any channel
- Slack — connect via bot token, agent joins your workspace
- Web Dashboard — built-in chat interface, no external app required
- OpenRouter (Claude, GPT-4, Gemini, Llama, Mistral, …)
- OpenAI, Anthropic, Groq, Ollama (self-hosted)
- Adjustable temperature per agent
- Each agent runs in an isolated container
- API keys and tokens are encrypted at rest (AES-256-GCM)
- Automatic subdomain + SSL per agent (
<name>.zeroonec.xyz) - Session auth with Google OAuth or email/password
| Free | Pro | Enterprise | |
|---|---|---|---|
| Agents | 1 | 5 | Unlimited |
| Channels (Telegram / Discord / Slack) | ✓ | ✓ | ✓ |
| Memory | ✓ | ✓ | ✓ |
| All AI providers | ✓ | ✓ | ✓ |
| Priority support | — | — | ✓ |
| Custom limits | — | — | ✓ |
Each agent runs on ZeroClaw — an open-source agent runtime built in Rust. It uses ~5MB RAM per instance and starts in under 10ms, which means ZeroOne can run many agents on a single server without the infrastructure cost ballooning.
Built with Potluck Labs