Google releases Omni Flash, a video generation model that improves consistency and real-world knowledge over its predecessor Veo

The Verge AI · May 23, 2026

AI Summary

•Omni Flash is now available in Google's AI video generation and editing platform, Flow. It can take a video upload plus a text prompt as input, and Google claims it incorporates more real-world knowledge when producing videos and keeps characters more consistent throughout a video compared to Veo.
•Users can edit videos with text-based prompts; Omni accepts and applies these edits better than Veo did, though results are inconsistent. The model can also add AI-generated elements to real videos, producing deepfakes that are often convincing enough to pass casual inspection despite occasional glitches (duplicated background figures, unnaturally manufactured sounds, uncanny visual tells).
•Video generation costs credits—15 to 40 credits per clip depending on length and inputs, with 40 credits per round of edits. The $20-per-month AI Pro plan includes 1,000 credits monthly; testing around 20 clips with edits consumed most of that allocation.

Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.