Cloud Workspace

Splatware

Splatware is a cloud workspace for creating, editing, sharing, commercializing, and exporting Gaussian Splatting models from images, video, and existing splat files.

Stage: Online Platform / Cloud TrainingNo-code training360 video workflowsEngine export guides

What It Does

Splatware is useful for users who want a web-based production workspace instead of assembling capture, training, editing, export, and documentation from separate projects.

The docs are also valuable even if you do not use the platform: their engine-integration guides point users toward current practical routes for Blender, Unity, and Unreal.

How To Use It In 3DGS

  • Prepare images, normal video, 360 video, or an existing model according to Splatware’s upload guidance.
  • Create a model in the cloud workspace and inspect quality before exporting.
  • Export PLY or .splat based on the target viewer, editor, or game engine.
  • Follow the matching integration guide for Blender, Unity, Unreal, or browser delivery.

Things To Watch

  • Cloud training quality still depends heavily on source coverage, lighting, motion blur, and camera path.
  • Output formats and plan features can change, so verify export requirements before building a client pipeline around it.
  • When using third-party engine plugins from Splatware docs, check the plugin’s own license and platform limitations separately.

Useful Input Rules

  • Splatware’s upload docs recommend enough visual coverage rather than minimal samples: at least 20 high-resolution images or 20 seconds of video is listed as a practical floor.
  • The docs call out 360 cameras as a strong input source, which makes Splatware relevant for room-scale, venue, and walkthrough workflows.
  • Existing PLY or SPLAT uploads can be useful when Splatware is part of a hosting/editing/export chain rather than the original trainer.

Why It Is Worth Listing

  • It combines a product workflow with documentation for surrounding production tasks, so users can go from capture guidance to engine export without leaving the ecosystem.
  • It is especially helpful for non-research users who need practical recipes rather than training-code parameters.