What is lufin?

Lufin is a modern self-hosted file sharing service built with React, Elysia, and Bun. It provides client-side end-to-end encryption using AES-GCM, password protection for shared files, and configurable data retention based on file size. The service offers rich previews for images, audio, video, archives, spreadsheets, text files, and PDFs directly in the browser. It supports multiple database backends including PostgreSQL, MongoDB, and SQLite, along with storage options such as local filesystem, S3-compatible services like Cloudflare R2, and local MinIO. Deployment uses Docker Compose with automatic HTTPS via Caddy, or manual installation from source. The frontend is fully static with no server-side rendering required. The interface is available in 26 languages. There is no user authentication system — files are shared via links stored in browser LocalStorage.

Should you self-host lufin?

lufin scores 45/100 on selfhost.directory (how scoring works) and ranks #62 of 86 in File Sharing. It has 234 GitHub stars.

There was no stable release in the past 12 months, so check the repository's recent activity before you rely on it.

Setup is moderate: expect some configuration beyond a single command. It is MIT-licensed and written mostly in Typescript.

Its closest alternative on selfhost.directory is copyparty (82/100, 47 153 stars). It scores higher overall, so compare the two before you pick. See lufin vs copyparty side by side.

Features

Automatic file renaming
Client-side Archive Generation
Client-side file previews
Client-side Image Compression
Configurable data retention
Delete at first download
Docker Compose Deployment
End-to-end encryption
File sharing via links
LocalStorage link management
Metadata stripping
Multi-language Support
Multiple database support
Password Protection

How to install lufin

Install lufin with Docker Compose

Docker Engine and Docker Compose installed

Environment variables

DOMAIN, EMAIL, BACKEND_PORT, CADDY_HTTP_PORT, AUTO_HTTPS, POSTGRES_USER, POSTGRES_PASSWORD, POSTGRES_DB, MONGODB_USERNAME, MONGODB_PASSWORD, MONGODB_DB_NAME, S3_ENDPOINT, S3_ACCESS_KEY, S3_SECRET_ACCESS_KEY, S3_BUCKET_NAME, S3_REGION, SQLITE_DB_DIR

Clone lufin to your machine & cd into project's root
Run ./configure.sh file and answer questions
- If you're running lufin as part of a larger platform or with other web services, or using Cloudflare, or want to manage TLS certificates yourself, you will need a reverse proxy running in front of it. Answer "no" when asked whether you want to enable Caddy automatic HTTPS. In all other cases, it's recommended that you answer "yes".
- Tip: if you want to keep automatic HTTPS but change the default binding to 80/443 ports, you can answer "yes" and then manually change ports in the generated .env file!
- Choose database and storage for lufin
- Databases:
  - PostgreSQL is recommended for most cases
  - SQLite is a good alternative for low-end machines
  - MongoDB can be used as your personal preference
- Storages:
  - Local uploads is fastest and recommended for most cases
  - Local Minio S3 can be beneficial for very high load due to caching
  - Remote S3 can be used for machines with small disk capacity
Run ./run.sh command with selected database and storage passed as arguments

Start:

./run.sh start

Stop:

./run.sh stop

Reload:

./run.sh reload

Stop everything and permanently delete everything, including databases and locally uploaded files:

./run.sh permadel

Recommended installation method

Install lufin with Build from Source

Environment variables

MONGODB_CONNECTION_STRING, POSTGRESQL_CONNECTION_STRING, SQLITE_DB_PATH, S3_ACCESS_KEY, S3_SECRET_ACCESS_KEY, S3_ENDPOINT, S3_BUCKET, S3_REGION, UPLOADS_DIR, CORS_ORIGIN, GRACE_PERIOD, VITE_API_URL, VITE_CONTACT_EMAIL, PORT

Clone repository to the server machine
Install frontend following steps in frontend/README.md -> Install
Install backend following steps in backend/README.md -> Install
Set up a regular job for cleaning up expired pages to automatically run bun /path/to/lufin/backend/src/jobs/cleanup-expired-pages.ts command
- One way is to use Cron which comes with most linux installations
  - add 0 * * * * /home/youruser/.bun/bin/bun --env-file=/path/to/lufin/backend/.env /path/to/lufin/backend/src/jobs/cleanup-expired-pages.ts to the crontab, see crontab.guru to adjust frequency
- Or use systemd timers with a config like contrib/systemd-lufin-cleanup-expired-pages.timer and contrib/systemd-lufin-cleanup-expired-pages.service
Set up a system daemon that will run backend (command bun start in the /path/to/lufin/backend directory)
- One way is to use systemd which comes with most linux installations with a config like contrib/systemd-lufin-backend.service
- Backend must be running under the same user who created backend/.env file, this file contains sensetive values and should not be readable by other users
- You should not run backend as the root user or as any other sudoer, create a separate linux user (e.g. lufin) and restrict its access to only lufin directory
- You can use the PORT environment variable to set the backend API port
Configure your reverse proxy by pointing url from VITE_API_URL (in frontend/.env) to the lufin backend (see Web server example configuration below)
- The proxy must accept websockets connections (e.g. Caddy handles it automatically, but for nginx you must add Upgrade and Connection headers)
- If you're getting HTTP 413 errors, increase request size limit (e.g Caddy does not set any limit by default, in nginx it's 1 MB and can be configured via client_max_body_size)