Self-Hosting Falcon Builder
Run the full workflow and agent builder on your own server. Your database, your file storage, your AI provider keys — nothing about how your workflows run leaves your infrastructure.
Cloud or self-hosted
Cloud at falconbuilder.dev is the fastest way to start — managed infrastructure, automatic scaling, and a free plan with no setup. Self-hosted trades that convenience for full control: it’s the same builder, running entirely on servers you own, for teams that need their data to stay on their own infrastructure or want to connect internal services Cloud can’t reach.
What you get
The Community edition is the full builder: visual workflows, AI- generated agents, knowledge bases with document embeddings, hosted forms and chat interfaces, a website chatbot widget, scheduled and webhook triggers, and every integration node. One workspace per instance, unlimited members — the first account to sign up owns it, everyone after joins by invitation.
What’s not in it: billing, affiliates, and the marketing site — there’s nothing to bill when it’s running on your own hardware.
What runs
A fixed set of services, started together with Docker Compose:
- web & worker — the application and the workflow execution engine (prebuilt images from GitHub Container Registry)
- postgres (with pgvector) — the database, including knowledge-base embeddings
- redis — job queues for the worker
- minio — bundled S3-compatible file storage (swap for your own bucket if you prefer)
- caddy — reverse proxy with automatic TLS; the only service that publishes ports
- scheduler — runs scheduled triggers, timeouts, and retention jobs
Quick start
Requirements: Docker Engine 24+ with Compose v2.20+, 2 vCPU / 4 GB RAM to start, and for a public server a DNS name with ports 80, 443, and 9000 reachable.
git clone https://github.com/NeoSky-AI/falcon-builder-self-hosted cd falcon-builder-self-hosted ./setup.sh # asks for your public URL, generates every secret into .env docker compose up -d
Open .env and set your SMTP server for invitations and password resets, and add an AI provider key (OpenAI is recommended for knowledge-base embeddings). On a laptop, the defaults give you a working instance at http://localhost:3000 with nothing else to configure.
Configuration guides
Everything lives in .env; the deployment repo has a guide for each service you connect:
- Sign-in — accounts, invitations, password reset, and Google or Microsoft sign-in with your own OAuth apps
- Email — SMTP for transactional mail and inbound email triggers
- File storage — the bundled MinIO, or your own S3, R2, and similar
- Upgrading — release notes and version-pinning for every update
Common Questions
Is self-hosting free?
Yes. The Community edition has no license fee — you run it on your own infrastructure and pay only for your server, storage, and whichever AI provider keys you connect.
Is Falcon Builder open source?
The self-hosted deployment stack — the Docker Compose file, setup script, and operator docs — is open source under the Apache 2.0 license on GitHub. The application runs from prebuilt container images published to GitHub Container Registry; the app’s own source isn’t published.
What’s different from the Cloud plans?
The Community edition is the same builder — workflows, agents, knowledge bases, hosted interfaces, integrations, scheduled and event triggers — with one workspace per instance and unlimited members. Billing, affiliates, and the marketing site don’t exist in it, since there’s nothing to bill: everything runs on infrastructure you already own.
Does it support Kubernetes?
Not yet — Docker Compose is the supported path today. The stack is a fixed set of services (web, worker, Postgres, Redis, object storage, a reverse proxy), so adapting the Compose file to your own orchestrator is straightforward if you need to.
Which AI providers can I use?
Bring your own API key for OpenAI, Anthropic, Google, Mistral, or OpenRouter, or point at a local model server with Ollama. Whatever you connect is billed directly by that provider — Falcon Builder doesn’t mark up or meter it.
Do I need to run my own database and file storage?
The Compose stack bundles Postgres (with pgvector) and MinIO for file storage, so a single `docker compose up -d` gives you a complete instance with no external accounts. You can swap either for your own Postgres or an S3-compatible bucket if you’d rather manage them separately.
How do I keep it updated?
Each release is a tag on the deployment repo pinning a specific image version, with an UPGRADING.md entry describing what changed. Updating is `git pull`, review the new entry, then `docker compose pull && docker compose up -d` — the migrate service brings your schema forward automatically.
Is there a workflow or execution limit?
No platform-enforced limit — you’re bounded only by your own server’s resources and whatever rate limits your AI provider sets.