Verified project record

elie222/rakazo

rakazo is an open-source platform for running persistent AI teammate bots that keep their own conversations, memory, routines, and history over time. Users bring their own model credentials and can self-host the stack, with sandboxed browser, terminal, file, and graphical desktop access for each bot.

Project overview

As an open-source alternative to Grok Bot, rakazo lets users choose both the model provider and sandbox provider for persistent AI teammates, and a one-script installer makes the self-hosted stack runnable without cloning the repository or installing Node.

Project type: MCP · AI Agent
Use cases: Automation
Deployment: Refer to project documentation
License: Apache-2.0

Best for

Users who want persistent AI teammate bots and are willing to bring their own model credentials, such as an OpenRouter API key.
Users who want control over model and sandbox choice, including self-hosted deployment with local Docker sandboxes.

Key capabilities

Bots have their own conversations, memory, routines, and history that persist over time.
Bots get browser, terminal, file, and graphical desktop access inside sandboxed computers, using local Docker or remote providers.
Bots can speak replies, accept dictated input, and take calls using user-supplied ElevenLabs, OpenAI, Cartesia, or Fish Audio keys.
Bots can delegate tasks to peer bots or short-lived subagents.
Users supply their own model credentials, such as OpenRouter or other supported providers, via Pi.
Integrations are available via Composio or Pipedream Connect, plus user-installed Treg endpoints, remote MCP servers, and OpenAPI tool documents.

Limitations and risks

rakazo is in beta.
Node.js 23.x and 25.x are not supported for development.
Desktop 'This computer' mode requires Docker Desktop, OrbStack, or Docker Engine to be installed.
Treg tool endpoints are usage-metered, and Treg's integration terms require a written agreement for hosted resale. Review these terms before relying on Treg endpoints in a hosted offering.
Running 'docker compose down -v' deletes all Postgres state; avoid it unless data loss is intended.

Getting started

No clone or Node install is required for published images. Create a rakazo directory, download install-images.sh, run 'bash install-images.sh' (the script downloads the Compose files, creates a .env with random secrets, and starts rakazo), then open http://127.0.0.1:5173, create an account, and connect a model. No coding is required.

Evidence and sources

GitHub project description: Open-source Grok Bot alternative. Choose your own model and sandbox.
README: Rakazo is an open-source platform for running persistent AI teammates.
README: It is available on the web, as an Electron desktop app, and through an Expo mobile app.
README: Bring your own model and computer provider, or run the complete stack locally.
README: Rakazo is in beta.

Source updated at: 2026-09-12T18:35:01Z

Open original GitHub project