Voxelo 3.0

Voxelo is a web‑based studio that transforms a single product video into interactive 3D models and AR experiences. By submitting just a short handheld clip, the service can deliver ready‑to‑use 3D assets that a retailer can embed into an online catalog. A free tier lets you host 100 “splats,” or you can create your initial 10 models directly from a video.

The workflow is broken down into Capture, Create, and Convert. In the Capture phase you record a product with a smartphone or any digital camera. The input is a normal video file, usually 30–60 seconds long, in common formats like MP4. The Create phase uses Voxelo's AI to analyze frames, reconstruct depth maps, and build a mesh that reproduces surface details and textures. You then use the platform's editor to tweak materials, changing color or glossiness, and embed JSON‑LD metadata that search engines can read. The Convert step packages the final model into formats such as OBJ, GLTF, or the proprietary “splat” format that can be embedded in websites. It also generates a place‑in‑space AR view that loads directly in a mobile browser, allowing users to place the item inside their living room or wardrobe.

Who benefits most is ecommerce operators that sell small, medium, or large products where visual detail matters, apparel, footwear, footwear, furniture, or any product that customers want to examine up close. Retailers who need to show stitching, fabric texture, or fit inside a room can use the AR view to reduce return rates and shorten the buying decision. Because the service offers a free hosting tier, it is attractive to small‑to‑mid‑market brands that want low‑cost prototyping. Larger enterprises may find the free 100‑splat cap restrictive, and the lack of detailed pricing information means they cannot evaluate scale costs before contacting sales.

Voxelo's description does not mention compatibility with particular e‑commerce platforms, nor does it state whether custom shaders, advanced lighting controls, or scripting for interactivity are available. There is also no detail on export limits per model, whether batch uploads are supported, or how frequently the AI can be re‑run on the same video. Finally, while it advertises JSON‑LD output, the site does not explain how compact or compliant the data is with existing schema.org vocabularies.

Key features