Your AI That Actually
Uses a Computer

ByteBot runs in a containerized desktop, performing real tasks: browsing the web, moving files, filling forms, and automating workflows -- just like a human would.

Get Started Free Watch Demo →

bytebot ~ task execution

The Problem

Automation tools only work within their own ecosystem

Zapier connects APIs. Selenium scripts web pages. RPA tools record clicks. But none of them can open a spreadsheet, research data online, fill out a form in a legacy app, and email the result -- all in one flow.

The Solution

ByteBot interacts with ANY desktop application, just like a human

ByteBot sees the screen, understands context, and takes action. It controls a real Ubuntu desktop environment -- clicking, typing, scrolling, browsing -- handling any application a person could use.

## Core Features

Everything You Need for Desktop Automation

A complete AI-powered desktop agent, running in Docker, controlled through a modern web UI.

Real Desktop Control

Full mouse, keyboard, and screen interaction inside a containerized Ubuntu environment. Not simulated -- real desktop actions.

Multi-LLM Support

Bring your own API key. Works with Anthropic Claude, OpenAI GPT-4, Google Gemini. Switch between providers on the fly.

Containerized Safety

Runs inside an isolated Docker container. The AI cannot access your host machine, files, or network unless you explicitly allow it.

VNC Access

Watch the AI work in real-time through a browser-based VNC viewer. See every click, every keystroke, every decision.

Task History

Every task is logged with full conversation history, screenshots, and execution timeline. Review, debug, and improve your workflows.

Modern Stack

NestJS backend, Next.js frontend, WebSocket real-time updates, PostgreSQL persistence. Clean TypeScript monorepo architecture.

View all features →

## Use Cases

What Can ByteBot Do?

Data Entry Automation

Extract data from emails, PDFs, or spreadsheets and enter it into any application -- CRMs, ERPs, legacy desktop software, web forms.

$ bytebot task "Read invoices from email and enter into QuickBooks"

Web Scraping & Research

Navigate websites, fill search forms, extract information, and compile results -- without writing a single Selenium script.

$ bytebot task "Search LinkedIn for AI engineers in NYC"

Report Generation

Open spreadsheets, analyze data, create charts, format documents, and export reports -- all autonomously.

$ bytebot task "Generate Q4 sales report from CSV data"

Application Testing

Test desktop and web applications by navigating UIs, filling forms, and verifying outputs -- no test scripts needed.

$ bytebot task "Test signup flow on staging.example.com"

## Technology Stack

Built With Open Source Leaders

Docker

Ubuntu

NestJS

Next.js

TypeScript

VNC noVNC

## Live Preview

Watch ByteBot Work

See a simulated task execution -- from command input to completion.

bytebot ~ task

Activities

14:32

Waiting for task...

Full interactive demo →

## Pricing

Simple, Transparent Pricing

Start free with the open-source version. Upgrade when you need managed infrastructure.

Open Source

Free

Self-hosted forever

✓Docker self-host

✓Community support

✓Single LLM provider

✓Basic task automation

✓MIT License

Clone Repository

Cloud

€29/mo

14-day free trial | €276/yr

✓Managed cloud instance

✓3 LLM providers

✓Unlimited tasks

✓Priority support

✓Task history & sharing

Get Started

Enterprise

€99/mo

14-day free trial | €948/yr

✓Dedicated instance

✓All LLM providers

✓API access & SSO

✓Custom workflows

✓SLA & on-premise

Contact Sales

See full comparison →

## FAQ

Frequently Asked Questions

What is ByteBot?

ByteBot is an autonomous AI agent that controls a containerized Ubuntu desktop. It can browse the web, manage files, fill forms, run terminal commands, and complete multi-step workflows -- just like a human sitting at a computer.

Is it safe? Can the AI access my computer?

ByteBot runs inside an isolated Docker container. The AI controls a sandboxed Ubuntu desktop, not your host machine. It cannot access your files, network, or system outside the container unless you explicitly mount volumes.

Which AI providers does it support?

ByteBot supports Anthropic Claude, OpenAI GPT-4, and Google Gemini. Configure your preferred provider's API key in the .env file and switch between them through the UI.

Do I need to pay for the open-source version?

No. The open-source version is completely free and MIT licensed. You only need a paid API key from your chosen AI provider (Anthropic, OpenAI, or Gemini). The Cloud and Enterprise tiers add managed infrastructure and premium features.

What are the system requirements?

Docker Desktop, 8GB+ RAM (16GB recommended), SSD storage, and an AI API key. ByteBot runs on macOS, Windows, and Linux -- anywhere Docker runs. The desktop environment inside Docker is always Ubuntu.

View all FAQ →

Ready to Automate Your Desktop?

Get started in under 5 minutes. Clone the repo, add your API key, and let ByteBot handle the rest.