Fireshare

Share your game clips, videos, or other media via unique links.





Live Demo  ·  Report a Bug

If Fireshare is useful to you, GitHub Sponsors is the best way to support development. It keeps the demo server running and helps me carve out time on weekends for new features and updates.

Key Features

• Share videos through unique links
• Public / private feeds (private is link-only)
• Video transcoding with CPU or GPU
• HDR to SDR tone mapping for washed-out HDR captures
• Password protected videos
• Game-based organization with cover art, including games you add by hand with your own artwork
• Folder sharing
• Mobile device support
• Uploads (optional, can be restricted)
• Video view counting
• Video cropping
• Video tags for improved search and categorization
• Open Graph metadata for rich link previews
• Notifications to Discord and others
• RSS feed for new public videos
• Multiple users with per-user permissions
• Shareable user profiles for uploads
• Two-factor authentication (TOTP authenticator apps)
• Login IP whitelisting
• Upload tokens for scripts and other tools

Supported Video Formats

Supported encodings: H.264 (AVC), H.265 (HEVC), AV1, VP9.

Files must be in a supported format because the original file is always served directly to the viewer. Transcoding (when enabled) only generates additional lower-quality versions for adaptive streaming; the original is never modified.

Navigation

• Installation
• Configuration
• Local Development
• Contributing
• Troubleshooting

Dashboard

Automatic Game Organization

Video Details

Public / Private Uploading

Installation

Fireshare is designed to run in Docker.

Required volume mounts:

Docker Compose

services:
  fireshare:
    container_name: fireshare
    image: shaneisrael/fireshare:latest-lite
    ports:
      - "8080:80"
    volumes:
      - /path/to/data:/data
      - /path/to/processed:/processed
      - /path/to/videos:/videos
      - /path/to/images:/images
    environment:
      - ADMIN_USERNAME=your-admin-username
      - ADMIN_PASSWORD=your-admin-password
      # A key for signing login cookies is generated on first start and kept in
      # /data, so there is nothing to set here. See docs/Security.md
      # The domain your instance is hosted at. e.x: demo.fireshare.net
      # this is required for opengraph to work correctly for shared links.
      - DOMAIN=
      # PUID/PGID: the user/group ID the container runs as. Files written to your
      # volumes (data, processed, videos, images) will be owned by this user. Set these to
      # match the owner of your host directories to avoid permission errors.
      # Run `id` on your host to find your UID and GID.
      - PUID=1000
      - PGID=1000
      # The port the web server (nginx) listens on inside the container (default: 80).
      # Leave this alone and change the host-side "ports" mapping above unless you run
      # with network_mode: host, where "ports" is ignored and this becomes the host port.
      # - FIRESHARE_PORT=80

Update the volume paths and credentials, then run:

docker-compose up -d

Open http://localhost:8080.

Docker Run

docker run --name fireshare \
  -v $(pwd)/fireshare/data:/data:rw \
  -v $(pwd)/fireshare/processed:/processed:rw \
  -v /path/to/my/videos:/videos:rw \
  -v /path/to/my/images:/images:rw \
  -p 8080:80 \
  -e ADMIN_USERNAME=your-admin-username \
  -e ADMIN_PASSWORD=your-admin-password \
  -e PUID=user-id-with-read-write-permissions \
  -e PGID=group-id-with-read-write-permissions \
  -d shaneisrael/fireshare:latest

Open http://localhost:8080.

Fireshare Lite

The fireshare:latest-lite image is a smaller alternative that uses the system-provided FFmpeg instead of the CUDA-enabled build included in the standard image. It is a good fit for most users who do not need GPU transcoding.

Use the lite image by appending -lite to your tag:

Note: Setting TRANSCODE_GPU=true has no effect in the lite image. GPU transcoding is permanently disabled regardless of environment variables.

Configuration

Security (IP Whitelist & Two-Factor Authentication)

Fireshare can restrict logins to a whitelist of IP addresses/CIDR ranges and supports TOTP two-factor authentication with any authenticator app. See Security.md for setup instructions.

Transcoding (Optional)

When enabled, Fireshare will create lower quality versions of your original supported file type videos. Your viewers can then choose to play your videos at lower qualities that their internet can handle. Fireshare will also attempt to automatically downgrade the quality of a viewer who is constantly buffering.

Transcoding is off by default. To enable it, set the following environment variables:

ENABLE_TRANSCODING=true
TRANSCODE_GPU=true   # optional, NVIDIA only

CPU transcoding works out of the box. For NVIDIA GPU transcoding you need three things: an NVIDIA GPU with NVENC support, the NVIDIA driver and NVIDIA Container Toolkit installed on the host, and the container started with GPU access (--gpus all or runtime: nvidia).

The image ships with NVIDIA_DRIVER_CAPABILITIES=compute,utility,video, which is what makes the NVIDIA Container Toolkit mount the NVENC libraries into the container. You do not need to set it yourself. If you do override it, it must include video (or be all), otherwise nvidia-smi will work inside the container but ffmpeg will have no NVENC encoder and Fireshare will fall back to CPU transcoding.

GPU requirements: NVIDIA GPU with NVENC support.

Docker / Docker Compose Setup

1. Install the NVIDIA driver and NVIDIA Container Toolkit on the host.
2. In docker-compose.yml, uncomment either runtime: nvidia or the deploy.resources.reservations.devices block (or pass --gpus all to docker run).
3. Set:

   ENABLE_TRANSCODING=true
   TRANSCODE_GPU=true
   

Unraid Setup

1. Install the "NVIDIA Driver" plugin from Apps / Community Applications.
2. Add to the Fireshare container environment:

   ENABLE_TRANSCODING=true
   TRANSCODE_GPU=true
   NVIDIA_DRIVER_CAPABILITIES=all
   

   NVIDIA_DRIVER_CAPABILITIES is optional on current images, which already default to compute,utility,video. It is required on older images, and harmless to keep.
3. Add --gpus=all to "Extra Parameters".

Encoder Selection

Fireshare automatically selects the best available encoder.

GPU mode (TRANSCODE_GPU=true):

CPU mode (TRANSCODE_GPU=false):

HDR Videos

HDR captures (10-bit BT.2020 with the PQ or HLG transfer, as NVIDIA and console recorders produce) play back flat and washed out in most browsers, and plain transcodes inherit the problem. Fireshare flags these videos and can make a tone-mapped SDR copy at the source resolution, which then plays as the Source quality and is what the lower qualities are made from. The original file is never modified, and the copy can be removed again from the video's editor.

Open an HDR video, choose Edit, and use the Dynamic Range row: pick an operator and apply it, or Compare the operators side by side on a few random frames of the video before choosing. Under Settings → Transcoding you can set the default operator and turn on automatic conversion of every HDR video (transcoding.tonemap_default and transcoding.auto_tonemap in config.json).

Auto picks bt.2390 when libplacebo can run on a GPU and hable otherwise. When a GPU is available every operator runs on it through libplacebo; without one, hable, mobius and reinhard run on the CPU. Each operator shows a GPU or CPU badge where it is chosen. Operators the server's ffmpeg cannot run are greyed out. Both images ship every operator; for bt.2390 to use the GPU the container needs graphics in NVIDIA_DRIVER_CAPABILITIES, which the full image sets by default.

Docker Environment Variables

See EnvironmentVariables.md for the full list of available environment variables.

Local Development

Requirements: Python 3, Node.js, and npm.

1. Clone the repo:

   git clone https://github.com/fireshare-app/fireshare.git
   

2. Start backend services from the project root:

   ./run_local.sh
   

3. Start the frontend:

   cd app/client
   npm install
   npm start
   

4. Open http://localhost:3000 and sign in with admin / admin.

Contributing

Contributions are welcome. For larger changes, open an issue first to align on scope.

1. Know how to code! I DO NOT WANT VIBE CODERS SUBMITTING PRs AI assisted coding is fine if you can read, understand, and validate AI code. Asking AI to review itself doesn't count as reading, understanding, and validating AI code.
2. Fork the repository
3. Create a branch from develop
4. Commit your changes
5. Rebase on latest develop
6. Open a pull request to develop

For issues and feature requests, visit the issue tracker.

Please DO NOT open a pull request for a feature or addition that was not previously discussed with me. Pull requests that do this will be automatically closed.

Database Changes

If you update models, create a migration and review it before opening a pull request.

Troubleshooting

See TROUBLESHOOTING.md for a full guide covering installation issues, playback problems, permission errors, transcoding, and more.

Star History

If you like the project, consider giving it a star. It helps increase visibility and supports continued development.