Deploy LibreFang

One command to deploy LibreFang Agent OS to Fly.io.
Free tier. Persistent storage. No credit card required.

Free forever 256 MB RAM 1 GB storage

Run this command

Open your terminal and paste. It will install Fly.io CLI, open your browser to log in, then deploy automatically.

$ curl -sL https://raw.githubusercontent.com/librefang/librefang/main/deploy/fly/deploy.sh | bash

Authorize in browser

A browser tab will open. Sign up or log in to Fly.io (GitHub OAuth supported). The script continues automatically after authorization.

Done — dashboard is live

Your LibreFang instance is deployed at https://librefang-xxxx.fly.dev. Add LLM API keys via the terminal prompts or later with flyctl secrets set.

⚡

Always On

No cold starts, no 15-minute sleep. Your agents run 24/7.

🔒

Persistent Data

1 GB volume for SQLite. Your data survives restarts and deploys.

🌐

HTTPS + WebSocket

TLS certificates auto-provisioned. WebSocket for real-time chat.

💰

Free Tier

Fly.io free tier: 3 shared VMs, 3 GB volumes. No credit card.

FAQ

Do I need a credit card?

No. Fly.io free tier does not require a credit card for basic usage.

What LLM providers work?

All 27 providers: Groq (free tier available), OpenAI, Anthropic, DeepSeek, Ollama, and more. Set API keys as secrets after deployment.

Can I connect Telegram / Discord / Slack?

Yes. Set your bot tokens as Fly secrets: flyctl secrets set TELEGRAM_BOT_TOKEN=your-token

How do I update?

Clone the repo and run flyctl deploy --app your-app-name. Or re-run the deploy script for a fresh instance.

Is 256 MB enough?

For 1-2 agents without channel bridges, yes. LibreFang idles at ~40 MB. For heavier workloads, upgrade to 512 MB ($3/month on Fly.io).