.webp)
The cost depends on the machine-hours used. 1 RH (Render Hour) = 1 hour of operation on a standard CPU server (~1,25$). GPU rates start at 3.75 RH/hour. You can estimate CPU rendering costs using our online calculator, but GPU costs are harder to predict accurately. Therefore, we recommend running a test render of 5-10 frames and asking our support team if you have any questions. If you are rendering a static image (single frame), you can reduce the resolution or the number of passes for the test.
Yes, upon registration, you receive 4 bonus RH for testing. Please note: test renders are saved with watermarks. To remove them, simply top up your balance with any amount.
The auto-farm (CPU/GPU) is a fully automated process: you upload your project via the website, and the system distributes the tasks across dozens of servers. Manual rental (AnyDesk) provides remote access to a specific high-end server (up to GPU H200 or 4090) for working with non-standard software or training neural networks.
We support 3ds Max, Blender, Cinema 4D, After Effects, Maya, Houdini, SketchUp, Unreal Engine, Twinmotion, Lumion, and other popular packages listed on our website. Supported engines include Corona, V-Ray, Cycles, Redshift, Octane, and more. If you need specific software or a plugin, contact support—we will likely be able to install it or add it to our development roadmap for the near future.
Yes, we provide powerful GPU servers for ML/AI tasks. Configurations with V100, A100, H100, and H200 graphics cards (up to 141 GB VRAM per card) are available. Rental is done manually with per-second billing.
We strictly maintain the confidentiality of your data. The very act of registering on our platform implies our Non-Disclosure Agreement (NDA).
In automatic mode, you can allocate up to 20 servers to a single task (up to 5 in demo mode). This speeds up the process exponentially: rendering on 10 servers for 1 hour costs exactly the same as on 1 server for 10 hours. If necessary, this server limit can be increased by contacting tech support.
Usually, this is not an error, but the "post-processing" stage—stitching frames into a final video or archiving the result. This process can take anywhere from a few minutes to an hour, depending on the sequence size. If the render has been stuck at 99% for an unusually long time, please contact our support team.
When creating a task, in the "Cameras" section, you can select multiple cameras by checking their boxes. Please note: the render settings (resolution, format, etc.) specified during task creation will apply equally to all selected cameras.
The complete list is available on the dedicated page on our website. Alternatively, click the "Create task" button in the top left corner of your dashboard and select your software—a list of supported versions, engines, and plugins will then be displayed.
Additional manual server rental is not required. By default, Blender (Cycles) rendering runs on our automated GPU farm. You simply need to create a standard task through the website following the instructions.