What is OtoDock?

OtoDock is a self-hosted platform that turns Claude Code, Codex, and local LLMs into a team of collaborative AI agents for your organization. Agents run in kernel-level sandboxes on your server, with optional remote machine pairing for full hardware access. The platform includes multi-tenant agent sharing with four modes, department organization with delegation rules, role-based access control at platform and agent levels, SSO/OIDC and two-factor authentication, scheduled tasks and webhook triggers, phone calls via Twilio or Asterisk/FreePBX, in-chat document editing, image generation and editing, a community catalog of ready-made agents and MCP tools, and per-user cost budgets. It runs via a single Docker Compose stack or bare-metal for development. The Fair Source license allows free self-hosting up to 5 users, with paid seat licenses for larger teams; each release becomes Apache 2.0 after two years.

Should you self-host OtoDock?

OtoDock scores 49/100 on selfhost.directory (how scoring works) and ranks #104 of 209 in AI Agents. It has 193 GitHub stars.

It shipped 10 stable releases in the past year. The latest, v1.9.0, came out on 25 October 2026. New versions land roughly every 6 days.

Setup is moderate: expect some configuration beyond a single command. An official Docker image is available. It is Other-licensed and written mostly in Python.

Its closest alternative on selfhost.directory is Orca (91/100, 108 446 stars). It scores higher overall, so compare the two before you pick. See OtoDock vs Orca side by side.

Features

App building and deployment
Department organization
Encrypted Credentials
Image generation and editing
In-chat document editing
Kernel-level sandbox
Multiple AI engines
Multi-tenant agent platform
Phone integration
Remote machine pairing
Role-based Access Control
Scheduled and triggered tasks
SSO/OIDC authentication
Usage and cost budgets
Web browsing

FAQ

Is my data on my server?

Yes. Data only goes to the AI providers you connect and services you integrate. The install's housekeeping (community catalog, release check, subscription check-in) carries none of your data.

What does the install need?

A Linux server or Apple Silicon Mac with Docker, 4 GB RAM, 13 GB disk, and sudo once on Ubuntu 24.04+.

Which AI accounts do I need?

A Claude subscription for Claude Code, a ChatGPT subscription for Codex, an API key, or a local model server. Each person connects their own or an admin shares a pool.