Verified, then logged.
Sun 22 Nov 2026 · 00:16 KST / 15:16 UTC한국어 | ENSearch /

Breaking01 / AI

Luma links video and sound generation with targeted editing

Luma described a workflow that combines sound and video generation with edits that preserve approved compositions.

A studio monitor shows vertical drama clips being edited, while a nearby smartphone plays the same imagined nighttime bus-stop scene with two distant figures seen from behind. AI-generated image

Luma outlined combined video and sound generation and follow-up editing in a post published on 14 Nov. The company said supported models generate dialogue and ambient sound alongside video.

According to Luma, the workspace brings video, image and audio work together. Its audio tools include voiceovers and sound effects. Audio-driven lip-sync adjusts characters’ mouth movements to speech.

Luma said Ray 3.2 is designed to assess its output and retry before presenting it. Multiple reference frames let users control movement and camera angles. It also supports EXR exports, an image format used for compositing.

Luma said its Uni-1 technology treats objects, text and backgrounds as separate editable layers. Users can replace products or wording while preserving an approved composition. The approach keeps the context of earlier decisions through revisions.

Luma described early video models as producing silent footage that needed separate sound work. Generating sound and pictures together lets teams review a complete audiovisual scene. The intended benefit is less manual work aligning sound with images.

The post did not disclose which models support which audio features or the prices for generation.