install# Selenix - AI-Powered Browser Automation
Version 4.0.1-beta.20 | Free | No installer required
Selenix is a free desktop application for browser automation, web scraping, and data extraction. It has a built-in AI assistant that can create automations from plain English — no programming knowledge needed.
Website: https://selenix.io Documentation: https://selenix.io/docs Download page: https://selenix.io/download
This ZIP contains portable builds of Selenix. There is no installer — just extract and run.
| Folder | Platform | How to run |
|---|---|---|
win-unpacked/ |
Windows 10/11 (64-bit, x64) | Run Selenix.exe from the folder |
mac/ |
macOS 10.15+ (Intel & Apple Silicon) | Run Selenix.app from the folder |
linux-unpacked/ |
Linux (64-bit) | Run the selenix binary, or use the .AppImage |
- RAM: 4 GB minimum, 8 GB recommended
- Disk: 200-300 MB free space
- Browser: Chrome v88+ (recommended) or Edge v88+. Must be installed on your system.
This is a beta release built in spare time. We do not currently have the budget to purchase code signing certificates for Windows or macOS.
On Windows: You will likely see a "Windows protected your PC" SmartScreen warning when running Selenix.exe for the first time. Click "More info" then "Run anyway" to proceed. This is because the executable is not signed, not because it contains anything harmful.
On macOS: You may see a warning that the app is from an "unidentified developer." Right-click (or Control-click) the app and select "Open" from the context menu, then click "Open" in the dialog. You only need to do this once.
- Extract the ZIP file for your platform
- Run Selenix (see table above)
- Create a free account at https://selenix.io/register (or log in if you already have one)
- Log in to the app using the same account you use on selenix.io
- Open the built-in browser and navigate to any website
- Use the AI assistant to describe what you want to automate, or record actions manually
All users get a free daily quota of AI messages included with their account. No credit card or payment is needed.
For full documentation and guides, visit https://selenix.io/docs
- AI Assistant — Describe what you want in plain English and the AI builds the automation for you. It can see your browser, read page content, take screenshots, and debug failures automatically.
- Auto-Run — Tell the AI what you want, and it generates commands, runs the test, checks if it worked, and fixes any problems — fully hands-off, up to 5 iterations.
- Web Scraping — Extract product listings, prices, contact info, job postings, news articles, or any structured data from websites. Export to CSV or JSON.
- 135+ Commands — Click, type, scroll, scrape, make HTTP requests, run JavaScript, send emails, upload files, and much more — all without writing code.
- Test Scheduling — Schedule your automations to run hourly, daily, weekly, or monthly.
- Browser Snapshots — Save and restore cookies, localStorage, and login sessions so you don't have to re-authenticate every run.
- Code Export — Export your automations to C#, Java, JavaScript, Python, or Ruby if you want to run them outside Selenix.
- Integrations — Send data to Zapier, Make, n8n, Power Automate, or any API using built-in HTTP and webhook commands.
- MCP Server — Connect Selenix to Claude Desktop via the Model Context Protocol for AI-driven automation from Claude's chat interface. Included in .zip and unpacked folder/resource/mcp-server/index.js OR install via
npm install -g @selenix/mcp-server. - Template Library — Pre-built automation templates for common tasks available at https://selenix.io/product/templates
- Account system: log in with your selenix.io account
- AI messaging tied to your account with free daily quota
- Production API endpoint configuration
- Registration moved to website (selenix.io/register)
New Features:
- Auto-Run Workflow — Fully automated test generation loop: describe what you want, and the AI generates commands, runs the test, verifies the result, and fixes any failures — up to 5 iterations with no manual intervention
- MCP Server (Claude Desktop Integration) — Connect Claude Desktop to Selenix via the Model Context Protocol. 18 automation tools accessible — screenshots, page HTML, test management, command execution, and more. Install via
npm install -g @selenix/mcp-server - Auto-Run Verification — After a test passes, the AI takes a screenshot and verifies the output actually matches your original request, catching issues like wrong pages, 404 errors, or incorrect data
- Targeted Fix Commands — AI can apply surgical fixes to specific commands by index instead of regenerating the entire test
Improvements:
- MCP bridge server starts after session is fully initialized, preventing premature tool calls
- Bridge client re-reads connection config on every call, so restarting Selenix does not require restarting Claude Desktop
- Command list now returns both camelCase identifiers and display names, preventing "Unknown command" errors
- Auto-Run status bar with color-coded states (blue=running, green=success, red=failed)
- MCP server bundled as a single file (1.4 MB) with no external dependencies
Bug Fixes:
- Fixed
run_testbridge endpoint hanging after test failure - Fixed
get_command_listreturning display names instead of camelCase identifiers - Fixed selector sanitization for AI-generated Playwright-style pseudo-selectors
New Features:
- AI Command Suggestion Cards — The AI Assistant now shows interactive command cards with "Add to Test" buttons, letting you insert suggested commands into your test with one click
- sendEmail Command — Send emails directly from your automation workflows
Improvements:
- AI Assistant system prompt updated to output machine-readable command blocks alongside natural language explanations
- Command name normalization handles case-insensitive matching for all 40+ commands
New Features:
- macOS DMG Installer — Native macOS disk image for streamlined installation on Mac
Improvements:
- Improved
scrapeCollectioncommand with better descriptions and support for @ attribute selectors - Enhanced
httpRequestcommand reliability and configuration
Bug Fixes:
- Fixed scheduler time calculation for accurate test scheduling
- Fixed XPath selectors containing @ symbols not being parsed correctly
- Fixed import feature not saving comments alongside test data
New Features:
- executePowerShell Command — Execute PowerShell scripts from within your tests, pass arguments including variables, and capture structured output
- Desktop Application — Selenix now runs as a standalone desktop app with system tray support, minimize-to-tray, and custom application icons
- License Management — Built-in license activation and validation system with encrypted license keys
Improvements:
- Application window shows minimize balloon notification on first minimize to tray
- Removed unnecessary console logs for cleaner production builds
- Improved EXE logger output for better debugging
Bug Fixes:
- Fixed CSV export for arrays producing malformed output
- Fixed PowerShell command not handling multiple arguments correctly
New Features:
- sendToAI Command — Leverage AI to generate, transform, or analyze data during test execution by processing existing variables or generating new content from instructions
- Smart Element Locator — Visual element selector with crosshair coordinates, element metadata display, and advanced selector tooltips
- AI Screenshot Analysis — Capture browser screenshots and send them to the AI Assistant for visual element identification and selector suggestions
- Import/Export Workflows — Import and export test workflows as JSON files for backup and sharing
Improvements:
typecommand now supports all keyboard keys and key combinations likeCtrl+Atypecommand now accepts variable interpolation in values- AI Assistant now sends chat history for contextual follow-up conversations
- Added intermediate status messages during AI processing
- Added stop button and keyboard arrow navigation to AI Assistant chat
- Adding a new test while a suite is selected now adds it to the current suite
Bug Fixes:
- Fixed
transformVariablenot working correctly with certain data types - Fixed import feature not loading saved test data properly
- Fixed
clickAtCoordinatestarget field not accepting user text input - Fixed AI instructions for screenshot-based locator detection
New Features:
- AI Assistant — Built-in AI chat interface in the Logger panel with continuous conversation, message editing, deletion, and clear history
- monitorElement Command — Watch elements for changes over time with configurable intervals and conditional stop values
- combineVariables Command — Merge multiple variables into arrays, objects, or strings
- inspectVariable Command — Debug and inspect variable contents, types, and structure during test execution
- createSnapshot / restoreSnapshot Commands — Save and restore variable state snapshots with JSON export and import
- Test Scheduling — Schedule tests to run automatically at configured times
Improvements:
- AI Assistant uses two-step processing: first gathers information, then returns structured results
- AI can run multiple special context commands in sequence for comprehensive analysis
Bug Fixes:
- Fixed array and object handling in
combineVariables - Fixed
getPageHTMLcontext command not returning content - Fixed AI Assistant displaying raw AI response in the log output
New Features:
- scrapeCollection Command — Extract arrays of similar elements from web pages using CSS or XPath selectors
- scrapeStructured Command — Extract complex structured data using JSON field mapping with support for nested selectors
- scrollAndWait Command — Automatically scroll pages and wait for dynamic content to load, with infinite scroll support
- transformVariable Command — Transform variable data with filtering, mapping, sorting, and format conversion
- downloadFiles Command — Download files from URLs directly within test execution
- exportToJSON / exportToCSV Commands — Export scraped data and variables to JSON or CSV files
- importFromJSON / importFromCSV Commands — Import data from external files into test variables
- httpRequest Command — Make HTTP requests with full configuration for headers, body, query parameters. Includes a visual request builder
- curlRequest Command — Execute cURL commands directly from tests for advanced API interactions
- File and Folder Upload — Upload single files, multiple files, or entire folders with MIME type detection
- Element Selection Tooltip — Visual tooltip when hovering over elements in the browser, showing selector information
Improvements:
- HTTP request UI with reorderable parameter lists and double-click configuration
- Variable interpolation support in HTTP request URLs, headers, body, and parameters
echocommand updated to handle objects and complex data types- Export commands support objects, strings, and arrays with automatic format detection
Bug Fixes:
- Fixed scroll command behavior on dynamic pages
- Fixed dialog box UI glitches during file operations
- Fixed JSON parsing for variable types in HTTP request body and parameters
- Fixed HTTP request save functionality
| Feature | Status | Expected |
|---|---|---|
| Chrome Driver Integration | Coming Soon | Q2 2026 |
| Cloud Execution | In Development | Q3 2026 |
| Template Marketplace | Planned | Q3 2026 |
- Documentation: https://selenix.io/docs
- Contact: https://selenix.io/support
- Account & billing: https://selenix.io/dashboard
Apache License 2.0 — see LICENSE for details.
Selenix is built on the foundation of Selenium IDE by the Selenium project.