High-performance native capture (<100ms latency), 100% offline Vision OCR, smart automated PII & sensitive redaction, multi-model AI vision integration (Claude, Gemini, GPT), and dedicated MCP Server for AI coding agents.
Mask passwords, credit cards & API keys automatically
Gemini, Claude, GPT-4o & OpenRouter
Engineered natively with Swift 6.0 to deliver silky 60fps performance without background resource hogging.
BerryShot supports dedicated capture modes: interactive region selection with pixel-magnifying loupe, clean window capture with alpha transparency, or batch capturing all open windows of an app simultaneously.
Inspect exact proportions and details before 1-click Copy, Save As, or Cloud Upload.
Seamlessly stitch long web pages, PDFs, and code files into a single high-resolution image.
Quickly pin screenshots on top, send to AI assistant, or copy OCR text with one click.
Built strictly with Swift 6 and AppKit/SwiftUI — zero Electron bloat, zero background lag.
Extract readable text from any screen coordinate using Apple Vision Framework. Works completely offline with zero telemetry and instant clipboard copying.
Automatically identifies and conceals credit cards (Luhn validated), passwords, emails, phone numbers, and API tokens before sharing.
High-frame-rate recording with simultaneous system audio and microphone capture. Pause, resume, and export compact HEVC MP4 video files.
Keep important references floating above all windows while writing code, comparing design mockups, or entering spreadsheets.
Sample exact HEX, RGB, or HSL color codes with a magnified pixel loupe and copy directly into your clipboard.
All credentials and API tokens are encrypted in the Apple Keychain Enclave. Transparent and source-available under PolyForm license.
Connect BerryShot with autonomous coding agents such as Claude Code or Codex. Your agent can request UI window captures, fetch view manifests, and generate test documentation securely via private stdio IPC.
# Add BerryShot MCP to Claude Code with one command:
claude mcp add berryshot -- /Applications/BerryShot.app/Contents/Helpers/BerryShotMCP
Trigger any capture mode instantly without breaking your current workflow.
Drag to select any area with high-precision coordinate & dimension loupe.
Select an isolated window or batch capture all windows of a running application.
Automatically scroll to capture entire websites, chats, and code repositories.
Copy edited screenshot or extracted Vision OCR text directly into clipboard.
Revert previous canvas shapes and markings safely with full undo history.
Instantly close the active capture overlay and return control to macOS.
Explore the authentic native UI built adhering to Apple Human Interface Guidelines.
Full-screen visual editor with shape tools, markers, and instant privacy blur.
Interactive response window for AI debugging, translation, and structured data output.
Dedicated selector for single or batch windows with aspect-ratio previewing.
Configure automated PII redaction rules and local stdio MCP agent servers.
Setup Claude 3.5, Gemini 2.0, GPT-4o, and customize prompt presets.
Auto-upload to Google Drive or multipart POST to your custom HTTP endpoint.
Download the official DMG release or explore the source code on GitHub. Completely free for noncommercial use.