Alent.Red sees your screen, moves your mouse, types on your keyboard, just like a human. More importantly: it learns from every interaction and remembers everything, forever. The longer it runs, the better it knows you.
Alent.Red watches your screen, moves the cursor, types keystrokes, and carries out complex workflows, wherever you run it: local, container, VPS, GPU or serverless.
The agent captures and interprets your screen in real time, recognizing text, buttons, images, menus, and UI elements with precision.
Pixel-perfect movement, left/right clicks, double-clicks, drag-and-drop, scrolling, exactly as if a human were at the controls.
Types text, triggers shortcuts, fills out forms, writes code with the same fluidity as a real user sitting at the keyboard.
Browser, office suite, terminal, code editor, business software: Alent.Red works with whatever is open on your screen.
Run it locally, in Docker, over SSH, on a $5 VPS, on a GPU cluster, or serverless (Daytona, Modal). One agent, your infrastructure, zero lock-in.
OpenAI, Claude, OpenRouter (200+ models), Nous Portal, or your custom endpoint. Plug in whichever model fits, with no vendor lock-in.
The agent retains your preferences, projects, and context, permanently. It learns from every interaction and improves session after session. No resets, no amnesia: the longer it runs, the better it knows you.
Compatible with the existing skills ecosystem and with the Model Context Protocol (MCP). Extend its capabilities through the open SKILL.md standard or by plugging in any MCP server. No need to reinvent the wheel.
Free for individuals and companies under $300K annual revenue: no subscription, no paywall, no usage caps. Above that threshold, a commercial enterprise license applies.
The agent learns from every interaction and remembers your context forever. The longer it runs, the better it knows you.
Tell Alent.Red what you need in plain English. It handles everything else.
Type a plain-text instruction: "Open Chrome, go to Gmail, and reply to my latest email"
AI vision captures your display, identifies every visual element, and plans the exact sequence of actions.
Alent.Red moves the cursor, clicks, types, and navigates the way you would, only faster.
Pick your OS, your infrastructure, and your model. The agent adapts: no account, no credit card, no strings attached.
Works on:
Supported LLM providers:
Because it controls your screen, Alent.Red is compatible with every app you already use.
"Open Gmail, read my unread emails, and draft a summary reply to my manager."
"Open the Q4 sales spreadsheet, build a chart from the revenue column, and export it as PDF."
"Find the top 5 Italian restaurants in Brooklyn and paste their addresses into a new doc."
"Open VS Code, create a new Python file, and write a web scraping script using BeautifulSoup."
"Go to Twitter, post my draft with the #AI hashtag, then like the first three replies."
"Sort the files on my desktop by type, rename the screenshots, and move them into folders."
Alent.Red is 100% free to use, for individuals and businesses alike: install it, run it, and put your daily tasks on autopilot with no subscription, no usage caps, and no credit card.
You get the full product from day one: AI vision, complete mouse and keyboard control, perpetual memory, and continuous learning. Nothing is held back.
And yes, you can build a real business on it: Alent.Red stays free as long as your activity generates less than $300K in annual revenue.
Above that threshold, you switch to a commercial enterprise license that keeps the project alive and funds the free tier for everyone else.
Ready to scale past $300K? Get in touch and switch to the enterprise license in minutes.
Yes, for the vast majority of users. Alent.Red is 100% free for individuals and for companies under $300,000 in annual revenue: no paid tier, no freemium, no usage limit, no credit card. Above that threshold, a commercial enterprise license is required. Get in touch.
Alent.Red uses AI vision to capture and understand your screen in real time. It then generates native mouse and keyboard commands through your OS system APIs (macOS, Windows). Think of it as a virtual user that can see and interact with your display.
Alent.Red runs on macOS 12+ and Windows 10+ (beta). It requires at least 8 GB of RAM and about 2 GB of disk space. A GPU is not required if you use a hosted model, but improves performance for local inference.
Join thousands of users who save hours every day with Alent.Red.