Adobe Firefly gets text-to-motion and layered video generation

AI · 5 min read

Adobe's Firefly update introduces text-to-motion generation that creates animated clips and motion templates from natural language prompts. The engine can output layered video with separate alpha, depth, and motion channels for editors to refine in Premiere and After Effects, instead of flat, baked footage.

Firefly's models are integrated into Creative Cloud so editors can iterate on motion style, timing, and camera behavior without leaving NLE timelines. Adobe also included a content-aware cleanup tool for generated footage to remove artifacts and match grain and lighting across layers.

For teams, Firefly now supports enterprise model governance, usage auditing, and asset provenance tracking so studios can comply with editorial and IP workflows. Creative directors can lock generation parameters and distribute approved motion presets to maintain brand consistency.