Björn Kindler has released RadianceKit 1.10, which lets users publish trained scenes by link or submit them to a community gallery.
RadianceKit is a native Mac app that trains gaussian splats on Apple silicon through Metal, with no cloud upload. Kindler builds it alone as Kindler Dev, near Munich. Other local Mac trainers include LuluSplat, which also trains splats on the Apple silicon GPU.
RadianceKit offers a guided Simple mode and an Expert mode with a three panel layout, classic or MCMC densification, and live loss curves. RadianceKit imports JPEG, HEIF, PNG and RAW photos, video, and 360 footage in equirectangular, cubemap, EAC and fisheye layouts. Sky masking reduces outdoor floaters, and a benchmark mode reports PSNR, SSIM and LPIPS. Exports cover PLY, compressed PLY, SPZ, glTF, .splat and SOG, plus orbit video and an interactive web viewer.
Version 1.8.1 added a scene editor in July, and version 1.8.10 added direct import of Insta360 X6 .insv files using the images from both lenses. The free iPhone companion, RadianceCap, stores camera pose, optics and LiDAR depth with every photo for training on the Mac.
RadianceKit 1.10 adds a RadianceKit Community entry under Export, which publishes a scene by link or submits it to a curated gallery on the RadianceKit Community site, optionally with its training settings attached as a community recipe. Align Scene moves and rotates the whole scene and sets its origin, and the placement is undoable and saved with the scene. Video exports no longer flicker on dense patterns and fabrics seen from mid distance, and renders show clean silhouettes without stair step edges. RadianceCap captures now train at about 3 megapixels, which Kindler says makes them about 20 to 30 percent faster.
RadianceKit is a free download with a three day trial of every feature. A one time $7.99 purchase unlocks the full version, including all export formats, quality presets, unlimited training and community publishing. RadianceKit requires macOS 26 or later on an M1 or later Mac.
RadianceKit 1.10 is available now on the Mac App Store.





