About screenpipe
Screenpipe is a local first memory layer for the desktop. It records screen content, app text, browser context and audio on macOS, Windows and Linux, so that the user and their AI tools can search what was seen or heard. Text is extracted through accessibility APIs with OCR as a fallback, and audio is transcribed locally with Whisper. Raw history stays on the device by default.
Users connect that history to AI tools such as Claude and Codex through MCP, so an assistant can draw on real work context. The newer Workflows app discovers recurring tasks across applications, documents the steps and highlights what could be automated, with source evidence for each step. The engine is source available, and local models can be run through Ollama or cloud models connected instead. The company is backed by Y Combinator.
It suits individuals and small teams who want searchable meeting and work history, personal recall and context for agents.
Pricing has four tiers. Free is $0 for one device with recent history. Basic is $21 per month, or $250 billed annually, for full searchable history and unlimited scheduled workflows. Business is $42 per seat per month, or $500 per seat annually, adding personal device sync and managed team seats. Enterprise is custom. Eligible students and researchers get 50% off the first month of an individual plan.