Self-host an open source LibreChat alternative you own
You self-host LibreChat to get the chat interface. Kortix is the open-source AI Operating System you self-host when you also want the agents, skills, memory, and connectors in one repo you own, with any model and your own keys.
Self-host the whole platform with three commands.
The self-host CLI, and the kortix.yaml that holds your agents and triggers.
curl -fsSL https://kortix.com/install | bashkortix self-host init --domain kortix.example.com --admin-email you@example.comkortix self-host start# kortix.yamlagents: support-triage: file: agents/support-triage.md # what it does, in plain markdown connectors: all # what it may reach# kortix.yamltriggers: - slug: daily-digest type: cron cron: '0 0 9 * * 1-5' prompt: Summarize yesterday's support tickets and open a change request with the digest.What self-hosting gives you here
Self-hosting LibreChat puts your model conversations into one interface you run yourself. Its own docs describe it as a self-hosted web app with no desktop installer, reached in the browser at localhost:3080 (LibreChat docs).
Self-hosting Kortix gives you a different object: one git repo that holds the company. Agents, skills, memory, tools, triggers and connectors are files in it (Kortix docs). You clone it with any git client, review each change as a diff, and keep the whole thing on hardware you control.
How the self-host install runs
The full platform runs as one Docker Compose stack, and the kortix self-host CLI installs, configures and updates it (self-hosting docs). Point A and AAAA records for the domain and the API subdomain at your server, open ports 80 and 443, then run the install. kortix self-host configure asks for your admin email, a sandbox provider key and the preview domain, and kortix self-host start applies the answers. kortix self-host doctor checks the result. The code panels on this page carry the commands.
Sessions run on a cloud sandbox provider, not on your server, so the instance stays reachable from the internet and the compute stays with that provider. You connect your own model keys in the web app after the first sign-in.
Two reachability modes cover most setups. Domain mode is the production path: Caddy terminates TLS on your domain and two replicas each of the API, the gateway and the frontend run behind it. Tunnel mode runs the web app on localhost:13737 and publishes only the API through a Cloudflare tunnel, which suits an evaluation on your own machine.
What you own after you install it
A Kortix project is a git repo with kortix.yaml at its root, and the work runs from that repo. Run it on Kortix Cloud, in your VPC, or on your own on-prem network. Self-host is free. Any model provider with your own keys. Secrets are encrypted at rest with a key per project.
Connectors are wired once for the whole project: 3,000+ apps in a click, plus MCP, OpenAPI, Postman, GraphQL and raw HTTP. Connector credentials are brokered server-side and never enter the machine. You set each connector to Allow, Ask or Block, down to the arguments of each call, and an Ask holds the call until a person approves it.
Work that lands through review
Each session gets its own isolated sandbox and its own branch. The agent commits, pushes the session branch and opens a change request against your default branch, and merge is default-deny for agents. A person reads the diff before anything reaches main. Thousands of agents in parallel on one config, each on its own cloud computer.
Which setup fits
Run LibreChat if a single chat interface over your conversations is the whole job, and you want it on your own machine.
Run Kortix if you want a repo the company owns, holding the agents, skills, memory, connectors and triggers, with sessions that return finished work and land through a change request a person reads.
self-hosted AI workspace guide · LibreChat alternatives · LibreChat vs Open WebUI
Self-hosting questions
No. The stack installs on a Linux server with a public IP and ports 80 and 443 open. Each session runs on a cloud sandbox provider, so the compute stays off your box.
Any model provider with your own keys. You add the key in the web app after the first sign-in.
It can be a Kortix-managed private repo, a repository in your own GitHub account, or a repo at any other git URL. On a self-hosted instance, a platform admin sets up managed repositories in the admin console.
A kortix-updater service pulls new images once a day, migrates the database and rolls the app services.
One repo you own, on your own hardware.
Start with Kortix Cloud, or run the whole stack yourself for free.
Open source · Any model, your keys · Self-host, VPC, or on-prem