OctaneRender 2027.1 Alpha 3 Is Out; 2026.5 Adds NVIDIA RTX Spark Support

Otoy released two versions of its GPU-based rendering engine, OctaneRender, in September 2026. The stable OctaneRender 2026.5, released on 13 September, adds native support for NVIDIA’s RTX Spark chip, developed for laptops and mobile workstations. On 24 September, Otoy released the third alpha build of the next major version, OctaneRender 2027.1. Alpha 3 includes quality improvements to the neural real-time rendering mode introduced in March, along with support for OpenPBR 1.2, SDF-to-mesh generation and OCIO 2.5.

OctaneRender is a physically based GPU rendering engine popular in architectural visualization, particularly among Cinema 4D, 3ds Max and Blender users. Known for its fast interactive previews, it is a popular choice for teams that render frequently during the design process.

OctaneRender 2026.5: RTX Spark and stability

According to the announcement on the Otoy forum, the headline feature in 2026.5 is the addition of native device binaries for NVIDIA RTX Spark. CG Channel describes RTX Spark as an unreleased “superchip” from NVIDIA, designed for laptops and mobile workstations. The release also includes:

• Support for OpenColorIO 2.5 configurations

• An increased limit on the number of mesh instances that can be loaded into a scene, provided there is enough VRAM

• Sample-count controls and continuous online training options for the neural radiance cache

• The --prevent-sleep command-line option, which prevents the computer from going to sleep until a script has finished

Fixes include a render hang when using photon mapping in network rendering, as well as dark artifacts in the caustics of objects with dispersion enabled. These fixes are significant for users working with caustics from glass, water and crystal objects in architectural scenes.

OctaneRender 2027.1 Alpha 3: neural rendering matures

The standout feature in 2027.1 is its neural real-time rendering mode, which, according to CG Channel, combines NVIDIA’s DLSS ray reconstruction technology with Otoy’s own research. The mode aims to deliver nearly noise-free interactive rendering at much lower sample counts.

Alpha 3 improves sampling, denoising and upscaling in this mode, aiming to provide a cleaner and more temporally stable interactive experience. Otoy lists the following improvements:

• More realistic glass transparency

• Fewer noise artifacts during sample accumulation

• Less flicker in specular reflections

• More consistent motion between frames

• Greater clarity in surface textures and lighting details

Other additions include an upgrade to MaterialX 1.39.6; an update to OpenPBR 1.2, which includes Specular Haze and retroreflection; SDF-to-mesh baking for use as proxy geometry in simulations; a UV atlas mode for high-quality displacement in UV-less projections; and a threefold increase in the maximum number of mesh instances.

Alpha 3 requires an NVIDIA Turing GPU (compute capability 7.5) or newer, and a driver version of at least 572. The real-time mode currently runs only on Windows and Linux; macOS support is planned for a later release.

Pricing and roadmap

According to CG Channel, OctaneRender is available with a Studio+ subscription for €23.95 per month or €239.88 per year, and the package includes plugins for 21 DCC applications. Free Prime editions, limited to a single GPU, are also available. Features Otoy plans to add in the future include 3D Gaussian Splat rendering and built-in PBR material generation.

Impact on archviz workflows

As the neural real-time mode matures, Octane users may be able to deliver smooth, low-noise previews in client presentations without switching to a separate real-time engine. Improvements to glass transparency and specular flicker are particularly relevant to glass-fronted buildings and glossy interior surfaces.

However, Otoy advises against using the alpha build for production work and warns that scenes saved with this version may not be compatible with future releases. For projects with delivery deadlines, sticking with the stable 2026.5 release is the safest option.

The significance of RTX Spark support for archviz is mobility. For visualization artists who want to edit a scene and render quickly during site visits, at a client’s office or at a trade show, GPU rendering on a powerful but slim laptop could be an attractive option. However, the performance and price of RTX Spark-based devices are not yet clear, so it is too early to say whether they will replace desktop workstations.

For offices using network rendering, the photon mapping fix and --prevent-sleep option in 2026.5 offer more immediate benefits. Render jobs queued overnight being interrupted because a computer goes to sleep is a familiar problem for many small offices.