Flock connects your local computers together into one private network for artificial intelligence. It works as a gateway between your desktop programs and your hardware. You use your own machines to process information. This keeps your data within your walls. You do not send your files or requests to big cloud companies.
The software acts as a central hub. It accepts requests from your code editors and AI tools. It then sends those requests to your available computers. This setup handles the work across different machines. You get faster responses. You control the costs. You manage the access.
You must visit the project page to download the latest version for your system.
Your main computer needs to run Windows 10 or Windows 11. You should have at least 8 gigabytes of memory. A solid-state drive helps the application start fast. If you plan to route work to other computers, ensure all machines share a stable network connection. A wired ethernet connection works better than wireless for heavy tasks.
- Visit the download link provided above.
- Find the section labeled Releases on the right side of the page.
- Click the most recent version number.
- Select the file ending in .exe for Windows.
- Save the file to your computer.
- Open your Downloads folder.
- Double-click the flock file to start the installation.
- Follow the prompts on the screen to finish the setup.
Flock works with standard software tools like Cursor, Aider, and Claude Code. These programs talk to external AI services. Flock replaces those external services with your local cluster.
- Open your AI editor or coding tool.
- Go to the settings menu.
- Look for the API settings section.
- Change the API address to match your local flock address.
- Enter your local flock access key if the software requests one.
- Save your changes.
The editor now sends requests to your machine instead of the internet.
You can add more computers to your cluster to increase power. This is useful for large tasks.
- Install flock on each additional machine.
- Start the flock service on every machine.
- Use the management dashboard to link machines by their network addresses.
- Set quotas for each user account to manage how much work happens.
The system automatically balances the work. If one computer runs out of memory, the system sends the task to the next available machine.
The software includes tools to track activity. You can see how many requests your tools send. You can also track the time each request takes to complete. Open your web browser and navigate to the local status page shown in the flock dashboard. This page shows charts and tables of your recent work.
If you see errors, check the logs on the status page. The logs show clear descriptions of what happened at each step. This helps you identify network issues or hardware limits.
Flock creates a bridge that you control. No information leaves your local network unless you tell it to. Because you host the gateway, you own the records of who used the system and what they requested. You can export these audit records at any time to review usage.
Do I need a server? No. You use your existing desktop or laptop computers.
What happens if my network goes down? The gateway stops processing requests. Once the network returns, the system reconnects and continues the work.
Does this cost money? The software is free to run. You only pay for the electricity used by your computers.
Can I use this for multiple users? Yes. The software supports different user keys. You can give each person a unique key and limit their total usage.
Does it work with existing AI models? Yes. It supports common industry standards. If your tool works with OpenAI or Anthropic, it works with flock.
Where do I see my API keys? Open the flock interface on your main machine. Click the account tab to generate or manage your keys.
How do I update the software? Visit the main download page, get the new version, and run the installer again. The program remembers your settings.
Keep your machines connected with high-speed cables. If you use multiple machines, place them on the same network switch. This reduces delays. Do not run heavy games or video editing software on the machine acting as the primary gateway. This ensures the gateway has enough power to handle incoming requests without lag. Close apps that you do not need while running your AI cluster. This frees up memory for better performance.