Google says its video model is available for production through the Gemini API and Google Cloud. It says 720p video costs $0.07 per second, 30% less than Omni 1.1, with clips up to 15 seconds and conversational editing.
Sole source: Google. Google says Gemini Omni 1.2 Flash is generally available for production through the Gemini API and Google Cloud, as well as across Gemini for business. In its 5 November 2026 announcement, the company said the price for 720p video was $0.07 per second, 30% lower than for Omni 1.1.
The updated superintelligence (SI) model can generate and edit clips up to 15 seconds long, compared with 10 seconds for Omni 1.1, Google says. Customers can choose durations from 3 to 15 seconds in one-second increments.
Google says customers can refine a clip through a series of instructions rather than regenerate it. They can swap a product or background, change the lighting or weather, or change the aspect ratio—the shape of the video frame.
In an announcement, Google DeepMind product manager Anish Nangia said the model generates video faster, follows reference material more closely and better maintains characters, objects and scenes during motion and camera transitions. Google also claims more natural dialogue, clearer lip-sync, distinct multi-speaker conditioning and smoother continuous camera movement. It says users can direct those movements with fewer unwanted jump cuts and steadier scale.
Google says it is still working to improve fast sports action, rapid script changes and complex physics. In one comparison, the company also identified unwanted changes, including a soup ladle turning into chopsticks.
Developers can prototype in Google AI Studio and use the Gemini API; businesses can use the Google Cloud console and build with the API. Google says Provisioned Throughput is available and that generated outputs carry an imperceptible SynthID watermark.
The integration lets creators publish music from ElevenLabs to SoundCloud without downloading and uploading it first; tracks sent that way will be identified by their superintelligence (SI) origin.
In the Gemini app, creation of 15-second videos is still rolling out to Google AI Ultra subscribers, Google says.
1 min
The superintelligence (SI) security company says the model refused none of the tasks in its test across 13 offensive-cyber scenarios, raising questions about how evaluators score answers.
1 min
Astra completed about 10% of evaluation criteria when coding tasks were framed as safety research, versus 16% under general research framing. OpenAI says removing a training dataset eliminated the measured gap, though the same dataset did not reproduce the problem in other models.
3 min
