Screenpipe SDK · for builders

Screen capture as a library.

Embed Screenpipe's capture engine in Electron, Swift, Tauri, and Node apps. Cross-platform screen + audio + OCR + accessibility capture for company-brain and agent products that need to know what happened across every app, including tools with no API.

Book a call

Screenpipe SDK blueprint — code snippet, capture primitives (screen, accessibility API, microphone, system audio, frame JPEG), trust metrics (300,000+ installs, 18K GitHub stars, 18 months in production), and use cases (training datasets, SOP generation, company brains, compliance replay).

SDK surfaces

One capture engine, four integration paths.

The SDK ships framework-specific surfaces instead of making every app build its own bridge. Pick the host runtime and keep the capture lifecycle consistent across products. Each app surface has a runnable example and smoke check in the repo.

Electron SDK

main + preload helpers

Use @screenpipe/sdk/electron to keep native capture in the main process while exposing a context-isolated renderer API for permissions, start, stop, snapshot, status, and reveal.

const { registerScreenpipeIpc } = require("@screenpipe/sdk/electron");

Swift SDK

Swift Package

Use ScreenpipeClient from a native macOS app with async/await, typed request models, JPEG snapshot decoding, and a testable transport layer.

import Screenpipe

Tauri SDK

frontend client + Rust plugin

Use @screenpipe/sdk/tauri with the Rust plugin to bridge Tauri commands into the same local capture lifecycle as Electron and Swift.

import { createScreenpipeTauriClient } from "@screenpipe/sdk/tauri";

Node SDK

native addon

Use Recorder directly for headless capture, eval harnesses, dataset generation, and custom desktop tooling.

import { Recorder } from "@screenpipe/sdk";

Building screen capture is a tar pit.

Different OS permissions. Audio routing. Accessibility APIs that change every year. DRM edge cases. macOS vs Windows vs Linux. Sleep/wake cycles eating CPU. PII.

Worse: most of the bugs only show up in the wild. Battery drain on a specific MacBook Air SKU. A Logitech webcam that crashes AVFoundation. A Windows update that quietly changes how UIAutomation reports off-screen elements. A Bluetooth headset that renames itself between sleep cycles. You can't catch any of that in CI — you catch it because a real user on that exact setup opened the app yesterday.

We've been shipping for nearly 2 years across 300K+ installs on every macOS and Windows version in active use, every Apple Silicon and Intel chip, plus Linux. That install base is our QA — every edge case has already been hit, reported, and fixed. License the capture engine and build what actually differentiates your product. Why not build it yourself?

What you get.

Drop-in capture engine

Bundle Screenpipe with your app. Use the Node core directly, or the Electron, Swift, and Tauri SDK layers for native desktop products.

Cross-platform, production-hardened

macOS (Apple Silicon + Intel), Windows x64 + ARM. Accessibility API for fast OCR-less capture. Audio routing handled. Battery-aware.

Local-first by design

Capture data stays on-device by default and is encrypted at rest with ChaCha20-Poly1305. Local PII detection runs before storage. Enterprise customers can opt into confidential-inference heavy-pass redaction.

Commercial embedding license

Priority support, SLA, roadmap input, dedicated engineer. White-label options available under OEM.

Private beta · enterprise only

Confidential-inference PII redaction.

For fintech, healthcare, insurance, and other regulated verticals, optional enterprise redaction runs Screenpipe's proprietary PII removal model/API inside a confidential-compute enclave (NVIDIA H200 TEE). Raw text is processed only inside the attested enclave and returned redacted, so Screenpipe and the cloud provider do not see it.

View ScreenLeak PII benchmark

8 PII categories benchmarked

Names, addresses, emails, phone numbers, URLs, dates, account numbers (credit cards, bank accounts), and secrets (passwords, API keys). F1 96% on PII-Masking-300k.

Optional confidential inference

Enterprise heavy-pass redaction for data you choose to send outside the device. Hosted by Screenpipe in TEE-backed GPU enclaves, with attestation proofs available.

Who embeds Screenpipe.

01Company-brain and agent startups — know what people are doing across Slack, browsers, docs, CRMs, IDEs, and legacy tools that do not expose an API. Capture the work graph first; build memory, automations, and agents on top. (e.g. Worktrace AI, building on Screenpipe in production.)
02Data companies — building computer-use datasets for labs. Every screen session becomes labeled trajectory data.
03Enterprise productivity tools — memory, meeting notes, CRM auto-capture, time tracking. Build the intelligence; skip the capture layer.
04Internal ops teams — embed in custom tools for workflow mapping, training data, knowledge capture across employee devices.

Want to embed Screenpipe?

30-minute call. Tell us about your product, we scope the embedding shape, and you leave with a proposal.

Book a call