Self-host Operant
Local-first multi-model AI agent for collaborative coding
- Stars
- 447
- Port
- 8000 (Local web UI and API server)
- Requires
- Python
What is Operant?
Operant is a self-hosted AI agent tool designed for complex programming tasks. It runs entirely on your local machine (macOS arm64 and Linux), providing a web UI, terminal interface (TUI), and a Tauri desktop wrapper. Multiple language models can be assigned distinct roles — planner, coder, reviewer — and collaborate through a structured message-passing system with scoped context. A hierarchical approval system separates clearly allowed actions from those needing human or model-assisted review, with hard denies that cannot be overridden. An experience and memory system captures verified knowledge from past tasks as versioned, scoped skills that can be recalled, reviewed, and shared across projects. The system is extensible via MCP and a local extension package format with six capability categories. All task state, conversations, and artifacts are stored in a local SQLite database. Model calls use your own OpenAI-compatible API endpoints and keys. The project is in beta, distributed as a macOS DMG and source builds, licensed under PolyForm Noncommercial 1.0.0 (source-available, not OSI-approved open source).
Should you self-host Operant?
Operant scores 39/100 on selfhost.directory (how scoring works) and ranks #170 of 209 in AI Agents. It has 447 GitHub stars.
It shipped 1 stable release in the past year. The latest, desktop-candidate-2026-09-28-08e65ea, came out on 28 September 2026.
Setup is advanced: plan for several services and some ongoing maintenance. It is Other-licensed and written mostly in Python.
Its closest alternative on selfhost.directory is Orca (91/100, 108 642 stars). It scores higher overall, so compare the two before you pick. See Operant vs Orca side by side.
Features
- Context compression and management
- File and session references
- Git-triggered workflows
- Local-first execution
- OpenAI-compatible model interface
- Plugin and extension system
- Skill discovery and management
Operant README
More in Code Management
pomeloAn open-source platform for running coding contests with sandboxed code execution, real-time scoreboards, and support for multiple programming languages.
pylonPylon IDE is a self-hosted, open-source web-based IDE forked from Cloud9 v2, featuring the ACE editor, built-in terminal, and plugin system on Node.js 22+.
QA-PilotQA-Pilot is a self-hosted, open-source chat app that helps developers understand GitHub code repositories using local or cloud LLMs with code graph visualization.
RoomoteSelf-hosted AI coding agent that runs in sandboxes, connects to chat apps, writes code, opens PRs with screenshots and live previews.
SeaGOATOpen-source semantic code search you self-host: vector embeddings plus ripgrep regex, local server with an HTTP API, and optional Ollama LLM enhancement.
sniboxSnibox is a self-hosted, open-source snippet manager built with Ruby on Rails and Vue.js for developers to collect, organize, and search code snippets.
snippet-boxSnippet Box is a self-hosted, open-source application for organizing and managing code snippets with built-in search, tagging, syntax highlighting, and Markdown support.
sturdySturdy is an open-source, self-hosted version control platform for real-time code collaboration, letting teams discover work-in-progress code instantly.
zedexzedex is a self-hosted, open-source proxy server for the Zed editor that caches extensions, serves releases, and provides edit prediction via OpenAI-compatible backends.
AtheosAtheos is a self-hosted, open-source web-based IDE running on PHP/Apache/Nginx with real-time collaboration, Git integration, and 40+ language support.
ByteStashSelf-hosted, open-source code snippet manager with language filters, search, sharing, an API, OIDC SSO and an MCP endpoint for AI assistants.
claude-context-localSelf-hosted, open-source MCP server that gives Claude semantic code search using PostgreSQL + pgvector and Ollama embeddings — 100% local with no API keys or cloud calls.
0 comments