Does OctaneRender have generative AI features?

Not inside the renderer yet. OctaneRender 2026, the current production release, uses AI to render faster and cleaner, not to generate content. Its main AI features are the Neural Radiance Cache, which cuts noise in indirect lighting, and path-traced rendering of Gaussian splats, which can come from AI generators. Generative features inside Octane, such as making a PBR material from a text prompt, are on OTOY's roadmap for Octane 2027. Generation today happens in OTOY Studio, OTOY's browser-based workspace for AI image, video, 3D and splat generation, which is a separate product.

What ships in Octane 2026 (production)

• Neural Radiance Cache (NRC). A neural network trained while you render, which reduces first-pixel noise in scenes lit mostly by indirect light. It arrived in 2026.1 on Windows and Linux, and 2026.4 (July 2026) added macOS and network rendering.
• Relightable Gaussian splats. Octane imports .PLY and .SPZ splats and path traces them, so they cast and receive shadows, show up in reflections and refractions, and take depth of field. The splats can come from real captures or from generators such as World Labs Marble. 2026.4 added SPZ v4.
• AI Light. A learned light-sampling method that has been in Octane since version 4.

What is in the Octane 2027.1 alpha

The 2027.1 alpha (Alpha 3 posted September 24, 2026, Windows and Linux only) adds a real-time neural viewport that combines OTOY's research with NVIDIA DLSS Ray Reconstruction, DLSS 4 and FSR 3 upscaling, and AI Light 2.0. OTOY has not announced a release date for 2027.1.

What is announced for 2027 but not shipped

OTOY's Octane 2027 roadmap (November 2025) lists:

• an AI Material Layer node that generates PBR materials from a text or photo prompt;
• rendering scenes directly to Gaussian splats ("Neural Objects");
• an anime and sketch renderer;
• OTOY Studio models inside the Octane node graph, for text-to-image, image-to-mesh and text-to-video.

None of these are in a production build as of October 2026.

How to use generative AI with Octane today

Generate in OTOY Studio and render in Octane:

1. Make the assets. Generate 3D meshes from an image with Meshy, Rodin, Hunyuan3D or Trellis, or make a Gaussian splat from photos, a video or a single image. World Labs Marble builds a whole environment as a splat. See how to make a Gaussian splat and image-to-3D models compared.
2. Export. Meshes download as GLB, and some models also give FBX, OBJ or USDZ. Splats download as PLY, including a version optimized for Octane, or as SPZ from Marble.
3. Render in Octane. Light and render them with everything else in your scene, in standalone Octane or a plugin such as Octane for Cinema 4D.
4. Bring renders back. Use the Octane render as the first frame or reference for a video model such as Kling, Seedance or Veo in OTOY Studio, so the camera and layout come from your scene.

OTOY Studio also makes textures, backplates, HDRI-style environment images and concept frames with image models such as Nano Banana Pro, GPT Image and Seedream. It is billed in credits, has a free plan, and does not require an OctaneRender license.

Last verified 2026-10-14. Prices and model lists come from OTOY Studio's billing configuration, exported 2026-10-20.