omnivore README

The project's README from github.com/omnivore-app/omnivore. As one markdown file: README.md.

Omnivore

GitHub Workflow Status Discord Mastodon Follow Twitter Follow GitHub

Omnivore Logo

Omnivore is a complete, open source read-it-later solution for people who like text.

[!IMPORTANT]
Omnivore is now a completely Self-Hosted application. The Cloud application was deprecated in November of 2024. The community still exists on Discord, and we endeavour to keep things updated and bug-fixes ongoing.

We built Omnivore because we love reading and we want it to be more social. Join us!

Every single part is fully open source! Fork it, extend it, or deploy it to your own server.

A guide for running a self hosted server can be found here

web-screenshot-listview

omnivore-discover

Join us on Discord! :speech_balloon:

We're building our community on Discord. Join us!

Shoutouts :tada:

Omnivore takes advantage of some great open source software:

Shoutouts: Community 🧑‍🤝‍🧑

And of course...

and many more who contributed in the past. If you feel like your name is missing from this list, please contact me (Podginator) via our Discord.

How to deploy to your own server

A guide for running a self hosted server can be found here

How to setup local development :computer:

The easiest way to get started with local development is to use docker compose up. This will start a postgres container, our web frontend, an API server, and our content fetching microservice.

Requirements for development

Omnivore is written in TypeScript and JavaScript.

Running the web and API services

1. Start docker compose

git clone https://github.com/omnivore-app/omnivore
cd omnivore/self-hosting/docker-compose/self-build
docker compose up

This will start postgres, initialize the database, and start the web and api services.

2. Open the browser

Open http://localhost:3000 and confirm Omnivore is running

3. Login with the test account

During database setup docker compose creates an account [email protected], password: demo_password.

Go to http://localhost:3000/ in your browser and choose Continue with Email to login.

Frontend Development

If you want to work on just the frontend of Omnivore you can run the backend services with docker compose and the frontend locally:

docker compose up api content-fetch
cd packages/web
cp .env.template .env.local
yarn dev

You will need to configure some values in the new .env.local file. These are the values for running the web service directly on your host machine and running api and content-fetch within docker:

NEXT_PUBLIC_BASE_URL=http://localhost:3000
NEXT_PUBLIC_HIGHLIGHTS_BASE_URL=http://localhost:3000
NEXT_PUBLIC_LOCAL_BASE_URL=http://localhost:3000
NEXT_PUBLIC_SERVER_BASE_URL=http://localhost:4000
NEXT_PUBLIC_LOCAL_SERVER_BASE_URL=http://localhost:4000

Running the puppeteer-parse service outside of Docker

To save pages you need to run the puppeteer-parse service.

1. Install and configure Chromium

brew install chromium --no-quarantine
export PUPPETEER_SKIP_CHROMIUM_DOWNLOAD=true
export CHROMIUM_PATH=`which chromium`

2. Navigate to the service directory, setup your env file, and install dependencies

cd packages/puppeteer-parse
cp .env.example .env
yarn

3. Start the service

yarn start

This will start the puppeteer-parse service on port 9090.

In your browser go to http://localhost:3000/home, click the Add Link button, and enter a URL such as https://blog.omnivore.work/p/getting-started-with-omnivore.

You should see a Chromium window open and navigate to your link. When the service is done fetching your content you will see it in your library.

License

Omnivore and our extensions to Readability.js are under the AGPL-3.0 license.

omnivore (omnivore-app/omnivore)

reader
Searching Database…

16.3K ·GitHub
What is this?Omnivore is an open source read-it-later app: you save articles, PDFs and newsletter issues into a personal reading queue, then read them with highlighting, notes, labels and search.
Is it just a website, or does it come with apps too?Both: there is a Next.js web app, a native iOS app, an Android app plus a PWA, and browser extensions for Chrome, Safari, Firefox and Edge. All of them talk to your own self-hosted server.
How do I get my own instance running?There is a guide for running a self-hosted server, and for local development the whole stack comes up with docker compose up — it starts Postgres, the web frontend, the API server and the content-fetching microservice, and creates a demo login for you.
What does a machine need before it will run?Docker for the compose stack. For working on the code you need Node.js v24.19 and Yarn (versions managed by Volta), plus Chromium, which the page-saving parser relies on.
Is there still a hosted version I can just sign up for?No — the cloud app was deprecated in November 2024, and Omnivore is now a completely self-hosted application. The community keeps updates and bug fixes going, so expect to run and maintain your own instance.
I read on my phone on the subway — will it work without a connection?Yes, offline support is one of the core features, so your saved queue stays readable without a connection.
If I modify it for my own server, do I have to publish my changes?Yes — Omnivore and its extensions to Readability.js are under AGPL-3.0, so network use counts as distribution and you must offer the source of your modified version.
How do articles actually get into my queue?Through the Add Link button in the web app, the browser extensions, or by emailing newsletter issues in. Saving runs through a puppeteer-parse microservice that fetches and renders the page with Chromium (port 9090 in dev).
Does it connect to Obsidian or Logseq?Yes to both — there is an official Logseq plugin and an official Obsidian plugin for getting your highlights and notes into those tools.
Can it read articles aloud to me?Text to speech is supported, but only in the iOS app for now.
Size
391.2 MB
Stars
16.3K
Age
4y
Commits · 90 Days