SheepIt Render Farm

Yes, SheepIt is completely free. The service relies on donations and ads to cover domain, server, and development costs. There are no hidden fees or charges for using the service.

Points determine the priority of your projects in the render queue. You earn points by contributing your computer's resources to render frames for others. You spend points when others render your frames. The calculation is based on render time, scaled to a reference machine (Intel Core i5-9600 for CPU and Nvidia RTX 3060 for GPU).

Yes, by default, your own projects will be rendered first. However, you can change this behavior in your account settings under the 'Options' tab if you prefer to prioritize other projects.

To participate, you need at least 8GB of RAM (with 2GB free), 2 CPU cores, and 40GB of free disk space. The recommended setup is 16GB of RAM. You also need to be able to run Blender 4.4. No additional software is required for Windows users, while Linux/OSX users need Java installed.

SheepIt stopped supporting AMD GPUs in January 2023 because frames rendered with HIP (AMD's rendering API) did not match the results from CUDA/Optix (Nvidia's APIs). This inconsistency prevented mixing renders from both GPU types. Additionally, there were very few AMD GPU contributors compared to Nvidia GPUs.

You can include a custom color configuration by placing a copy of your local OCIO folder in your project folder, renamed to 'OCIO'. Ensure the folder contains a file named 'config.ocio'. To reduce file size, remove unnecessary looks and LUTs, but make sure to update the 'config.ocio' file accordingly to avoid errors.

No, SheepIt does not allow NSFW or adult content projects. Since the platform does not enforce a minimum age requirement, it assumes all users are under 18, and such content is prohibited.

SheepIt automatically detects the best frame-splitting method during project upload. The two methods are 'chessboard' (splits frames into smaller tiles with a border pattern) and 'layers' (splits frames into layers with reduced samples, then recomposes them). The 'chessboard' method does not support compositing, while the 'layers' method does but requires a consistent sample count.