Google has quietly replaced Veo 3.1 with Gemini Omni Flash for video generation within its consumer Gemini application, a change that occurred around May 19, 2026. While Veo 3.1 remains available via API for higher-quality 4K output, the consumer-facing model now has less transparent usage limits and a reduced consistency in editing over multiple turns. Users are advised to adjust their prompting strategies for Omni Flash, focusing on scene meaning and style rather than technical camera details, and to manually verify outputs due to observed limitations in character and object consistency, text rendering, and lip-syncing. AI
IMPACT This shift impacts user experience and content creation workflows for Gemini video generation, highlighting current limitations in AI video consistency and control.
RANK_REASON The article discusses a change in the underlying model used for a consumer product, detailing its capabilities and limitations, rather than a new model release from a frontier lab.
- Demis Hassabis
- Engadget
- Gemini
- Gemini API
- Gemini Omni Flash
- Google DeepMind
- InVideo
- Medium
- Tamzidul Haque
- Veo 3.1
- Vertex AI
- X
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
