Everything you need to know about ByteBot. Can't find an answer? Contact us.
ByteBot is an autonomous AI agent that controls a containerized Ubuntu desktop environment. It can see the screen, move the mouse, type on the keyboard, browse the web, manage files, and complete multi-step workflows -- all without human intervention. It runs inside Docker and is controlled through a modern web interface.
Most automation tools work through APIs (Zapier), browser scripting (Selenium), or recorded clicks (RPA). ByteBot is different because it interacts with a real desktop environment like a human user -- it sees the screen, understands context through AI vision, and operates any application. This means it can handle tasks that span multiple applications, legacy software, and scenarios that traditional tools cannot handle.
Anything you can do on a desktop: browse the web, open applications, fill out forms, write documents, manage files, run terminal commands, create spreadsheets, send emails, and complete multi-step workflows. Common use cases include data entry automation, web scraping, report generation, and application testing.
Yes. The core ByteBot agent is MIT licensed and fully open source. You can self-host, modify, and distribute it freely. The Cloud and Enterprise tiers add managed infrastructure, premium features, and priority support on top of the open-source core.
Docker Desktop, 8GB+ RAM (16GB recommended), SSD storage, and an AI API key. ByteBot runs on macOS, Windows, and Linux -- anywhere Docker runs. An NVIDIA GPU is optional but recommended for running local AI models via Ollama.
Four commands:
Then access the UI at http://localhost:9992 and the desktop VNC at http://localhost:6080.
Anywhere Docker runs: macOS, Windows, and Linux. The desktop environment inside Docker is always Ubuntu regardless of your host OS.
Pull the latest changes and restart the containers:
ByteBot runs inside an isolated Docker container. The AI controls a sandboxed Ubuntu desktop, not your host machine. It cannot access your files or system outside the container unless you explicitly mount volumes. If anything goes wrong, you can simply destroy and recreate the container.
The VNC, UI, and API ports are designed for local use only. Never expose them to the public internet without proper authentication. If you need remote access, use a VPN or SSH tunnel. Also, always change the default PostgreSQL password in your .env file before deploying.
For self-hosted deployments, all data stays on your machine in Docker volumes and the PostgreSQL database. Nothing is sent to external servers except your prompts to the AI provider you choose (Anthropic, OpenAI, or Gemini). For Cloud/Enterprise tiers, data is stored on managed infrastructure.
ByteBot supports Anthropic Claude, OpenAI GPT-4, Google Gemini, and local Ollama models. Configure your preferred provider's API key in the .env file and switch between them through the UI.
If you use Anthropic, OpenAI, or Gemini, you need a paid API key from that provider. If you want a fully free setup, use Ollama -- it runs AI models locally on your machine with no API key or charges. Note that local models require more RAM and a capable GPU for best performance.
Yes. ByteBot supports switching between providers. Configure all the keys in your .env file and select the provider per task through the UI. The Cloud tier supports 3 simultaneous providers and the Enterprise tier supports all providers.
Yes. The open-source version is completely free and MIT licensed. You only pay for the AI provider API key (or use Ollama for free). There are no hidden costs, usage limits, or time restrictions.
For Cloud and Enterprise tiers, we accept PayPal (paypal.me/zelvi999), bank transfer (IBAN: AL-C0100002334708), and cryptocurrency via BSC (BEP-20). Contact elvizekaj@zedigital.tech for billing questions.
Yes. You can upgrade from Open Source to Cloud or Enterprise at any time. Downgrades take effect at the end of your current billing cycle. Contact support for plan changes.
Ensure Docker Desktop is running and has at least 8GB of memory allocated. Check that no other services are using ports 9992, 6080, 3001, or 5432. Run docker-compose logs to see detailed error messages.
Verify your API key is correctly set in the .env file. Check that the selected provider is accessible from your network. Look at the agent logs with docker-compose logs bytebot-agent for error details. Ensure your API key has sufficient credits/quota.
The Ubuntu desktop may need a few seconds to fully initialize after container startup. Wait 10-15 seconds and refresh. If the issue persists, restart the desktop container with docker-compose restart bytebot-desktop.
Reach out to us directly or open an issue on GitHub.