Web Hive: Designed with an intuitive interface leveraging state-of-the-art frameworks, this platform simplifies interactions with AI agents, delivering a smooth and efficient user experience.

Comprehensive LLM Integration: Our platform supports a variety of leading-edge Large Language Models (LLMs), such as Gemini, OpenAI, Azure AI, Anthropic, DeepSeek, and Ollama, with plans to continuously expand and include additional models in the future.

Seamless Browser Connectivity: Integrate your preferred browser effortlessly, eliminating repetitive login hurdles and simplifying authentication processes. High-quality screen recording is also included for enhanced functionality.

Uninterrupted Sessions: Keep your browser environment active between tasks, enabling seamless access to your interaction history and maintaining the operational context of AI activities.

Setup Options

Method 1: Install Locally

Refer to the Quickstart Guide or proceed with the steps outlined below to begin.

Requires Python 3.11 or later.

To begin, we suggest using uv to set up the Python environment.

uv venv --python 3.11

and activate it using:

source .venv/bin/activate

Install the required dependencies:

uv pip install -r requirements.txt

Next, install Playwright:

playwright install

Method 2: Docker Setup

1. Requirements:

   • Ensure Docker and Docker Compose are installed on your system
   • Git is needed to clone the repository

2. Setup:

   # Clone the repository
   git clone https://github.com/browser-use/web-ui.git
   cd web-ui
   
   # Copy and configure environment variables
   cp .env.example .env
   # Edit .env with your preferred text editor and add your API keys

3. Launch Using Docker:

   # Build and start the container with default settings (browser closes after AI tasks)
   docker compose up --build
   
   # Or run with persistent browser (browser stays open between AI tasks)
   CHROME_PERSISTENT_SESSION=true docker compose up --build

4. Open the Application:

   • Web Hive: http://localhost:7788
   • Use VNC Viewer (to observe browser interactions): http://localhost:6080/vnc.html

The default VNC password is set to "vncpassword." You can modify it by updating the VNC_PASSWORD environment variable in your .env file.

Usage

Local Setup

1. Copy the .env.example file to .env and configure your environment variables, such as API keys for the LLM:
   cp .env.example .env
2. Start the Application:

   python webhive.py --ip 127.0.0.1 --port 7788

3. Web Hive options:

• --ip: Specifies the IP address Web Hive binds to. Default: 127.0.0.1.
• --port: Specifies the port Web Hive binds to. Default: 7788.
• --theme: Defines the user interface theme. Default: Ocean.
  • Default: A balanced, standard design.
  • Soft: Muted colors for a relaxed aesthetic.
  • Monochrome: Grayscale for simplicity and focus.
  • Glass: A modern, semi-transparent look.
  • Origin: A retro-inspired, classic appearance.
  • Citrus: Bright and fresh colors with a vibrant feel.
  • Ocean (default): A calming, blue ocean-inspired design.
• --dark-mode: Activates dark mode for the interface.

3. Access Web Hive: Launch your web browser and go to http://127.0.0.1:7788.
4. Using Your Own Browser (Optional):
   • Configure the following environment variables:

     • CHROME_PATH: Set this to the path of your browser's executable.

     • CHROME_USER_DATA: Set this to your browser's user data directory.

     • Windows

        CHROME_PATH="C:\Program Files\Google\Chrome\Application\chrome.exe"
        CHROME_USER_DATA="C:\Users\YourUsername\AppData\Local\Google\Chrome\User Data"

       Note: On Windows systems, replace YourUsername with your actual Windows username in the paths.

     • Mac

        CHROME_PATH="/Applications/Google Chrome.app/Contents/MacOS/Google Chrome"
        CHROME_USER_DATA="~/Library/Application Support/Google/Chrome/Profile 1"

   • Close all Chrome windows.
   • Open Web Hive in a non-chrome browser, such as Firefox or Edge. This step is crucial as the persistent browser context will utilize Chrome's data when running the agent.
   • In the Browser Settings, enable the "Use Own Browser" option.
5. Keep Browser Open (Optional):
   • Add CHROME_PERSISTENT_SESSION=true to your .env file.

Docker Setup

1. Environment Variables:

   • All configurations are managed via the .env file.
   • Key environment variables include:

     # LLM API Keys
     OPENAI_API_KEY=your_api_key_here
     ANTHROPIC_API_KEY=your_api_key_here
     GOOGLE_API_KEY=your_api_key_here
     
     # Browser Settings
     CHROME_PERSISTENT_SESSION=true   # Enable to keep the browser open between AI tasks
     RESOLUTION=1920x1080x24         # Specify resolution in the format: WIDTHxHEIGHTxDEPTH
     RESOLUTION_WIDTH=1920           # Set the screen width in pixels
     RESOLUTION_HEIGHT=1080           # Set the screen height in pixels
     
     # VNC Settings
     VNC_PASSWORD=your_vnc_password  # Optional, defaults to "vncpassword"
     

2. Browser Persistence Modes:

   • Default Mode (CHROME_PERSISTENT_SESSION=false):
     • The browser starts and shuts down with each AI task.
     • Ensures a clean slate for every interaction.
     • Optimized for lower resource consumption.

• Persistent Mode (CHROME_PERSISTENT_SESSION=true):
• Keeps the browser open between AI tasks.
• Retains history and session state.
• Enables viewing of previous AI interactions.
• Can be configured in the .env file or set as an environment variable when starting the container.

3. Viewing Browser Interactions:

   • Open the noVNC viewer by navigating to http://localhost:6080/vnc.html.
   • Enter the VNC password (default: "vncpassword" or the value set in the VNC_PASSWORD environment variable).
   • You can now observe all browser interactions in real time.

4. Managing the Container:

   # Launch with persistent browser session
   CHROME_PERSISTENT_SESSION=true docker compose up -d  
   
   # Launch in default mode (browser resets after each task)
   docker compose up -d  
   
   # Monitor real-time logs
   docker compose logs -f  
   
   # Shut down and remove the container
   docker compose down  
   

Changelog

• 2025/01/26: Thanks to @tiagonascimento31 . Now Web Hive can combine with DeepSeek-r1 for enhanced AI capabilities!