What is screenpipe?

Overview

screenpipe is an open source application (MIT license) that continuously captures your screen and audio, creating a searchable, AI-powered memory of everything you do on your computer. It runs in the background, extracting text from your screen using OCR and transcribing audio using Whisper — all locally on your device.

With over 16,700 stars on GitHub and 1,300+ forks, screenpipe is the most popular open source screen memory tool available. It is commonly used as an alternative to Rewind.ai (now Limitless), Microsoft Recall, Granola, and Otter.ai.

Unlike cloud-based alternatives, screenpipe stores 100% of your data on your own device. You can use fully local AI models (via Ollama) so that no data ever leaves your machine, or connect to cloud AI models like ChatGPT, Claude, or Gemini for more powerful search and analysis.

How it works

1. Screen capture: screenpipe records all your monitors simultaneously at configurable intervals using platform-native APIs (CoreGraphics on macOS, DXGI on Windows).
2. OCR text extraction: Every captured frame is processed with optical character recognition to extract all visible text — from any application, browser, document, or chat window.
3. Audio transcription: System audio and microphone input are transcribed in real-time using OpenAI's Whisper model, running locally on your device.
4. Local storage: All extracted text, transcriptions, and compressed screenshots are stored in a local SQLite database on your device.
5. AI search: Search through your entire computer history using natural language. Ask questions like "What did we discuss in the standup yesterday?" or "Find the API key I saw in that Slack message last week."

Key features

Screen recording with OCR

screenpipe captures all connected monitors simultaneously and extracts text from every frame using OCR. It records which application was active, the window title, and browser URLs when available. This means you can search for text you saw in any application — emails, documents, chat messages, code editors, browsers, or anything else on your screen.

Audio transcription

All audio from your system (meetings, videos, music with lyrics) and microphone (your voice) is transcribed using OpenAI Whisper running locally. Speaker identification helps distinguish between different people in conversations. This works with any audio source — Zoom, Google Meet, Teams, or any other application.

AI-powered search

Search your entire screen and audio history using natural language. Filter results by date, application, window title, or browser URL. Results include screenshots and audio clips alongside extracted text, so you can see exactly what was on screen at any moment.

Timeline view

A visual timeline lets you scroll through your entire screen history like a DVR. Click any moment to see the full screenshot and extracted text. Play back audio from any time period.

Plugin system (Pipes)

screenpipe has an extensible plugin architecture. Popular plugins include automatic Obsidian note syncing, Notion integration, meeting notes with action items, and AI-generated daily journals. Developers can build custom plugins using TypeScript/JavaScript.

MCP server for AI assistants

screenpipe runs as a Model Context Protocol (MCP) server, allowing AI assistants like Claude Desktop, Cursor, and VS Code to directly query your screen history. This gives your AI assistant memory of everything you've done on your computer.

Developer API

A full REST API runs on localhost, providing endpoints for searching screen content, audio transcriptions, and frames. A JavaScript/TypeScript SDK is available. Developers can build custom applications on top of their screen data.

Apple Intelligence integration

On supported Macs, screenpipe can use Apple Intelligence for on-device AI processing — generating daily summaries, action items, and reminders with zero cloud dependency and zero cost.

Platform support

Minimum requirements: 8 GB RAM recommended. Approximately 5–10 GB of disk space per month of continuous recording. CPU usage is typically 5–15% on modern hardware.

Privacy and security

• 100% local by default: All screen recordings, audio, OCR text, and transcriptions are stored on your device in a local SQLite database. Nothing is sent to external servers.
• Open source: The entire codebase is MIT licensed and publicly auditable on GitHub. You can verify exactly what the software does.
• Local AI support: Use Ollama or any local AI model so that no data ever leaves your machine — not even search queries.
• No account required: The core application works without creating any account or signing in.
• You own your data: Export, delete, or back up your data at any time. The database is a standard SQLite file.
• Optional encrypted sync: If you choose to sync between devices, data is end-to-end encrypted with zero-knowledge encryption.

How screenpipe compares to alternatives

For detailed comparisons, see screenpipe vs alternatives.

Use cases

Meeting transcription and notes

screenpipe automatically transcribes all meetings across Zoom, Google Meet, Microsoft Teams, or any other platform. After a meeting, search for any topic discussed or use the meeting assistant plugin to generate structured notes with action items. Unlike dedicated meeting tools like Otter.ai or Granola, screenpipe captures your screen too — so you can find that slide, document, or chat message that was shared during the call.

AI coding assistant context

Give Cursor, Claude Code, Cline, or Continue access to your screen history via MCP. Your AI coding assistant can see what documentation you were reading, what errors appeared on screen, and what you discussed in meetings — making code suggestions dramatically more relevant.

Personal knowledge management

Automatically capture everything you read and see into a searchable archive. The Obsidian plugin syncs your screen activity into your knowledge base. Never lose track of that article, email, or conversation again.

ADHD and focus support

For people who frequently lose track of what they were doing, screenpipe provides a complete timeline of your day. Search for "what was I working on before lunch?" or find that tab you closed hours ago. The timeline view lets you visually scroll through your entire day.

Remote work and async collaboration

For remote teams, screenpipe provides automatic documentation of every meeting and work session. Share AI-generated summaries with teammates who couldn't attend. All processing happens locally — no sensitive work data sent to third-party services.

See more at screenpipe use cases.

Integrations

• AI coding assistants: Cursor, Claude Code, Cline, Continue, OpenCode, Gemini CLI
• AI chat assistants: ChatGPT (via MCP), Claude Desktop (via MCP), any MCP-compatible client
• Note-taking: Obsidian, Notion
• Local AI: Ollama, any OpenAI-compatible model server

Pricing

screenpipe uses a one-time payment model for the desktop application (not a subscription):

• macOS and Windows: One-time purchase (dynamic pricing, typically $300–$400). Includes lifetime access and all future updates.
• Linux: $200 supporter license (build from source).
• Open source CLI: Free forever (MIT license).

Optional: screenpipe Pro subscription ($29/month or $228/year) for cloud sync between devices with end-to-end encryption.

See current pricing and download →

Frequently asked questions

Is screenpipe free?

The core engine and CLI are free and open source (MIT license). The desktop app with the graphical interface requires a one-time purchase. There is no recurring subscription for the core app.

Does screenpipe send my data to the cloud?

No. All data is stored locally on your device by default. If you use cloud AI models (like ChatGPT or Claude), only your search queries are sent to those services. You can use fully local AI models via Ollama for complete privacy.

How much disk space does screenpipe use?

Approximately 5–10 GB per month of continuous recording, depending on screen activity. You can configure retention periods to automatically delete old data.

Does screenpipe slow down my computer?

screenpipe is designed to be lightweight. Typical CPU usage is 5–15% on modern hardware. You can adjust capture intervals and OCR frequency to reduce resource usage.

Can I use screenpipe with ChatGPT, Claude, or Cursor?

Yes. screenpipe runs as an MCP (Model Context Protocol) server, allowing Claude Desktop, Cursor, and other AI assistants to directly query your screen history. For ChatGPT, you can use the screenpipe API to provide context.

What makes screenpipe different from Microsoft Recall?

screenpipe is open source, cross-platform (macOS, Windows, Linux), includes audio transcription, has a plugin system and developer API, and lets you choose any AI model. Microsoft Recall is closed source, Windows-only, screen-only (no audio), and has faced privacy controversies.

Can I record multiple monitors?

Yes. screenpipe captures all connected monitors simultaneously. This is a key advantage over alternatives like Rewind/Limitless which only capture the active window.

Is there a mobile app?

Not currently. screenpipe is a desktop application for macOS, Windows, and Linux.