Gemini Omni 1.2 Flash Release Notes

Google has officially released Gemini Omni 1.2 Flash, a model update focused on video generation capabilities, cost efficiency, and improved consistency. The model is now generally available across the Gemini API, Google Cloud, and various Google consumer applications.

Key Technical Improvements

Gemini Omni 1.2 Flash introduces several functional upgrades for video production workflows:

• Extended Duration: Users can now generate and edit video clips natively up to 15 seconds, with the ability to specify duration in one-second increments starting from 3 seconds.
• Enhanced Reference Adherence: The model demonstrates better consistency when using character, product, or style images as references. This includes maintaining fidelity for faces, garments, and typography across different camera angles and motion sequences.
• Multi-turn Conversational Editing: The model supports iterative refinements. Users can modify specific elements—such as backgrounds, lighting, or weather—in subsequent turns without needing to regenerate the entire clip from scratch.
• Cinematic Control: New capabilities allow for complex, continuous camera movements, such as FPV drone flythroughs and long zooms, with reduced instances of unwanted jump cuts.
• Improved Audio and Lip-Sync: The update includes more fluid dialogue generation and better multi-speaker conditioning, intended for use in professional video creation tools.

Pricing and Availability

Google has reduced the price of the model by 30% compared to Gemini Omni 1.1, with the service now priced at $0.07 per second for 720p video generation. The model is accessible via:

• Developers: Available through the Gemini API and Google AI Studio, with support for Provisioned Throughput.
• Enterprise: Accessible via the Google Cloud console and the Gemini enterprise agent platform.
• Consumer Apps: Rolling out to Google Vids, the Gemini app (for Google AI Ultra subscribers), and Google Ads.

Community and Industry Reception

Initial feedback from the developer community has highlighted significant confusion regarding Google's product naming conventions and the staggered rollout schedule. Users have reported difficulty determining which models are available to specific subscription tiers, with some noting discrepancies between the text-to-speech audio and written documentation regarding pricing updates.

Some power users expressed skepticism regarding the practical utility of current AI video models for precise creative direction. As one user noted:

"I just don't know how video is going to make it, long-term. You have to go through so many iterations to get something with even a semblance of what you were going for. I would rather have to wait 5-10 minutes for a clip that actually does what I tell it to than get something useless in 30 seconds."

Others pointed out that while the technical advancements in consistency are welcome, the fragmented nature of Google's AI ecosystem—ranging from "Flash" to "Experimental" labels—remains a barrier to adoption for those seeking reliable, long-term production tools. Despite these concerns, the focus on ad-creation workflows suggests that Google is positioning the model as a core component of its broader advertising and enterprise media strategy.