Screenpipe

Screenpipe is a source-available desktop application that records and indexes screen activity, application text, meetings, and audio locally. It creates a searchable history of a user's work and allows selected context to be shared with Claude, Codex, Cursor, ChatGPT, and other MCP-compatible AI assistants.

What Is Screenpipe?

Screenpipe is a work-context capture tool for macOS, Windows, and Linux. It continuously captures information from a user's screen and microphone, then processes that information locally so it can be searched and used in AI-assisted workflows.

The application can extract visible text using operating-system accessibility data, with OCR available as a fallback. Audio can also be transcribed locally, allowing users to search previous meetings, conversations, and other recorded activity.

Screenpipe includes an MCP server that lets compatible AI assistants access context selected by the user. This can help an assistant understand what the user previously viewed, discussed, or worked on without requiring the user to manually copy and paste all of that information into a prompt.

Captured content remains on the user's device by default. Optional cloud features are available for users who want capabilities such as cross-device synchronization, team workflows, connectors, exports, or additional AI capacity.

Key Features

Local Screen and Audio Capture

Screenpipe can continuously capture screen activity and microphone audio on a user's computer. The captured information is stored locally by default and becomes part of a searchable work history.

Searchable Work History

Users can search previously viewed application content, screen text, meetings, and audio recordings. This can help recover information from previous work sessions without manually organizing every interaction into notes.

Accessibility Text and OCR

Screenpipe uses accessibility data to extract text from supported applications. When accessibility text is unavailable, OCR can be used to identify text displayed on the screen.

Local Audio Transcription

Recorded audio can be transcribed locally. This makes previous meetings and spoken conversations searchable without requiring all captured audio to be sent to a cloud transcription provider.

MCP Server for AI Assistants

Screenpipe provides an MCP server that can connect captured work context to compatible tools such as Claude, Codex, Cursor, and ChatGPT. Users can choose which information to make available instead of automatically sharing their complete activity history.

Recurring Workflows

Screenpipe can run scheduled and recurring workflows using information from captured activity. These workflows can be used to summarize work, organize information, or support other repeated tasks.

Cross-Device Sync

Eligible paid plans support syncing context across multiple devices. This allows users to access captured work information from more than one supported computer.

Team Workflows

Paid plans include features intended for teams that need to use captured context across shared workflows. Cloud-based team features are optional and separate from the application's default local storage model.

Screenpipe Use Cases

Screenpipe can be used to:

• Search previously viewed screen content
• Find information from past meetings and conversations
• Give AI assistants selected context through MCP
• Recover details from previous work sessions
• Create summaries from captured activity
• Run recurring workflows using work history
• Search locally transcribed audio
• Sync work context across eligible devices
• Support shared workflows across teams
• Reduce the need to manually copy information into AI prompts

How Screenpipe Works

A user installs Screenpipe on a supported macOS, Windows, or Linux computer and grants the operating-system permissions required for screen or microphone capture.

Screenpipe then captures selected screen activity, application text, meetings, and audio. The application processes and indexes this information so the user can search it later. Accessibility text is used where available, while OCR can provide a fallback for text displayed on the screen.

Users can connect Screenpipe to an MCP-compatible AI assistant and choose relevant context from their captured history. The assistant can then use that information to answer questions or complete tasks with a better understanding of the user's previous activity.

Captured content stays on the device by default. Information is transmitted externally only when the user enables cloud features, synchronization, connectors, exports, or other workflows that require selected data to leave the device.

Screenpipe Pricing

Screenpipe uses a Free + Paid pricing model.

The free version starts at $0 and provides local capture without requiring a credit card. Paid plans currently start at $21 per month when billed annually.

Paid plans add features such as full history, additional AI capacity, cross-device synchronization, and team workflows. The exact features and usage limits depend on the selected plan.

Limitations

Screenpipe requires operating-system permissions to capture screen activity and microphone audio. Users who do not grant the corresponding permissions will not be able to use those capture features.

Continuous recording can consume local storage. The amount of storage required depends on the user's activity, capture settings, and retention period.

Although captured content remains local by default, optional cloud AI features, synchronization, connectors, exports, and team workflows may transmit selected information when enabled. Users handling sensitive information should review these settings before activating cloud-connected features.

Screenpipe is source-available under the Screenpipe Commercial License. Source-available software is not necessarily the same as software distributed under a standard open-source license.

Who Is Screenpipe For?

Screenpipe is designed for developers, researchers, founders, knowledge workers, and teams that want a searchable record of their computer activity.

It is especially useful for people who regularly work with AI assistants and want to provide relevant context without repeatedly explaining previous tasks or manually copying information from applications, meetings, and documents.

Screenpipe may also be useful for users who want local-first screen and audio capture while retaining the option to enable cross-device and team features through a paid plan.