Dependency-aware uploads
Textures, caches, and add-ons are tracked automatically, so a project that renders locally renders the same way on the farm.
Rendering built to scale
OctaRender is a cutting-edge distributed rendering platform that gives artists and creators unparalleled speed and efficiency. Leverage our global network of nodes to render complex 3D scenes in a fraction of the time, all with simplicity in mind.
Scale from a single GPU to hundreds mid-job, with no queue and no minimum spend. Every node on the marketplace is browsable and hand-picked, with live pricing and verified performance scores, so there is never a black-box assignment.
From your Blender project to finished frames in four steps.
Install the Blender add-on, open your project, and set your render options. Textures, caches, and add-ons are tracked automatically.
Browse available GPU nodes with transparent, live pricing and hand-pick the hardware for your job.
Track node activation and rendering progress in real time from the project dashboard.
Download finished frames from the Blender add-on or the web viewer, individually or by folder.
Textures, caches, and add-ons are tracked automatically, so a project that renders locally renders the same way on the farm.
Watch node activation and render progress in real time, with visibility into what each node is doing on your job.
Browse the GPU marketplace and choose nodes by specs, region, and live price after you sign in, with no blind hardware assignment.
Multi-layer EXR output with per-File-Output-node passes, previewable frame by frame in the web viewer.
Stop an individual node or requeue a single frame without restarting the whole job.
Pause a running job and resume it later. Your queued and finished frames stay exactly where you left them.
Watch a full render job run from upload to finished frames.
Yes, render nodes are billed only while they have work. The platform checks for new tasks every couple of seconds and starts shutting a node down as soon as there is nothing left to render, so idle time is not billed.
You only pay for successful compute time. If a node goes offline mid-job, billing for that node stops immediately and the incomplete frame goes back into the queue to be picked up by another node, so you do not lose the work.
Yes. The upload step packages your project's dependencies, including textures, VDB simulation volumes, scripting assets, and custom add-ons, into a secure archive, so the render environment matches what you use locally.
Yes, open the project viewer in the web portal to watch frames come in and download the ones that are already finished, without interrupting nodes that are still working.
Need help or want to share your work? Join our Discord community