English | 简体中文 | 日本語

[ screenpipe ]

AI app store powered by 24/7 desktop history

open source | 100% local | dev friendly | 24/7 screen, mic recording

[ recording reality, one pixel at a time ]

 https://github.com/user-attachments/assets/628c6c01-a580-4b21-bce9-3e7b186914a4 https://github.com/user-attachments/assets/973ee8e5-5240-4d36-83fe-d38c53efe6a9 --- *news* 🔥 - [2025/03] introducing [screenpipe operator](https://github.com/mediar-ai/screenpipe/tree/main/pipes/desktop-to-table): playwright but for your desktop. low level API, 100% accurate, not pixel based unreliable agents like openai/anthropic - [2025/02] we're throwing an [hackathon](https://www.nosu.io/hackathons/screenpipe), $12k in cash prizes, 28 Feb - [2025/01] we're partnering with Different AI to bring you [financial automations based on your screen](https://github.com/different-ai/hypr-v0) and [drop-in replacement for granola within obsidian](https://github.com/different-ai/file-organizer-2000) - [2024/12] pipe store stripe integration: devs build cool shit - few lines of JS and make passive income (available Reddit agent, LinkedIn agent, Timeline ...) - [2024/11] [screenpipe is number 1 github trending repo (again)](https://x.com/louis030195/status/1859628763425931479) - [2024/10] screenpipe has been backed by [Founders, Inc](https://f.inc/) - [2024/09] [screenpipe is number 1 github trending repo & on hackernews!](https://x.com/louis030195/status/1840859691754344483) - [2024/08] anyone can now [create, share, install pipes](https://docs.screenpi.pe/plugins) (plugins) from the app interface based on a github repo/dir - [2024/08] we're running bounties! contribute to screenpipe & make money, [check issues](https://github.com/mediar-ai/screenpipe/issues) - [2024/08] we released Apple & Windows Native OCR. - [2024/07] **we just launched the desktop app! [Download now!](https://screenpi.pe)** --- # how it works? - we record everything 24/7, 100% locally, uses 10% CPU, 4 GB ram, 15 gb/m - we index it into an api - dev build ai apps w user's full context, desktop native, nextjs, publish, monetize # why? - ai models are commoditized - ai is as good as its context - the most valuable context is all contained in your screen ## get started macos, linux: ```bash curl -fsSL get.screenpi.pe/cli | sh ``` or on windows ```bash iwr get.screenpi.pe/cli.ps1 | iex ``` then ```bash screenpipe ``` make sure to allow permissions on macos (screen, mic) - [get the desktop app](https://screenpi.pe/) - [docs & build from source](https://docs.screenpi.pe/getting-started) ## create plugins ```bash bunx --bun @screenpipe/dev@latest pipe create ``` screenpipe has a plugin system called "pipe" which lets you create desktop app in nextjs in a sandboxed environment within our Rust code, [read more](https://docs.screenpi.pe/plugins) you can then publish these to our store and make money: ```bash cd foo bunx --bun @screenpipe/dev@latest pipe register --name foo [--paid --price 50] # subscription bun run build bunx --bun @screenpipe/dev@latest pipe publish --name foo ``` ## community - [template to build screenpipe-powered desktop native app using Tauri](https://github.com/LorenzoBloedow/screenpipe-tauri-template-dev) - [template to build screenpipe-powered desktop native app using Electron](https://github.com/neo773/screenpipe-electron) - [community projects](https://www.sprint.dev/hackathons/screenpipe) ## star history  ## contributing contributions are welcome! if you'd like to contribute, please read [CONTRIBUTING.md](CONTRIBUTING.md).