Screenpipe logoscreenpipe
GitHubMIT

Screen recording that understands content

Traditional screen recorders create video files. screenpipe creates searchable, AI-powered memory. Find any moment with natural language.

📧
🌐
💬
📹
📝
🖥️
capturing everything

See it in action

Screen recordings are useless

Recording your screen doesn't help if you can't find anything.

01

Hours of video with no way to search

02

Scrubbing through footage to find one moment

03

No text search - just visual scanning

04

Large files that fill up your drive

05

Recordings are write-only - never reviewed

Searchable screen recording

screenpipe records your screen, extracts text with OCR, transcribes audio, and indexes everything. Search in natural language, find any moment instantly.

OCR extraction

Every word on screen is extracted and indexed. Search for any text you've seen.

Audio transcription

All audio is transcribed automatically. Search meetings, videos, voice notes.

Natural language search

Ask 'that API documentation I read yesterday' - not exact keyword matching.

AI-powered answers

Don't just find moments - ask questions and get answers from your recordings.

How it works

1

Record everything

screenpipe captures your screen and audio continuously with minimal resource usage.

2

Automatic indexing

OCR extracts text, speech-to-text transcribes audio, everything gets indexed.

3

Search and find

Natural language search across all recordings. Jump directly to any moment.

Key benefits

Find any screen moment in seconds
Search spoken words, not just text
AI answers questions from your recordings
Efficient storage with compression
Private and local - nothing leaves your machine

Frequently asked questions

Start recording intelligently

Screen recording that actually helps you find things.