The open source Rewind alternative
Everything Rewind does - 24/7 screen recording, AI search, instant recall - but open source and private. Your data stays on your device.
Quick answer
Looking for an open source Rewind alternative? screenpipe offers 24/7 screen recording and AI search with a local-first, open-source architecture.
See it in action
Why look for a Rewind alternative?
Rewind is powerful, but it comes with limitations.
Your screen data is processed on Rewind's servers
Mac-only, no Linux or Windows support
Closed source - you can't verify what happens to your data
No API access for developers
screenpipe does the same thing, but open source
screenpipe captures your screen 24/7 and lets you search with AI. Core capture and search can run locally, with optional cloud AI, sync, exports, connectors, and team workflows scoped separately.
Open source
MIT licensed. Inspect the code, contribute, or fork it.
Privacy controls
Use local-only capture when you need device-local storage and clear boundaries before enabling sync, cloud AI, exports, connectors, or team workflows.
Cross-platform
Works on macOS, Windows, and Linux.
Developer API
Full API access to build your own integrations and automations.
How it works
Download and install
Get screenpipe from GitHub or our website. One-click install on Mac, Windows, or Linux.
It runs in the background
screenpipe captures your screen and audio with minimal resource usage.
Search anything
Ask natural language questions like 'what was that website I visited yesterday?' and get answers.
Key benefits
Frequently asked questions
Yes, screenpipe is open source under the MIT license. You can inspect the code, contribute, or fork it. The desktop app requires a license for full features.
screenpipe offers the same core features as Rewind: 24/7 screen recording, AI-powered search, and instant recall. Additionally, screenpipe is cross-platform (Mac, Windows, Linux), has a developer API, and keeps all data local. Rewind is Mac-only and processes data on their servers.
No, screenpipe is designed to be lightweight. It uses hardware-accelerated encoding and typically uses less than 1% CPU. Storage is optimized with compression, using about 5-10GB per day of recordings.
Yes. All processing happens locally on your device. Your screen recordings, extracted text, and audio transcriptions never leave your computer. You can verify this yourself since the code is open source.