- Google has officially announced “Gemini Omni 1.2 Flash,” the latest version of its video generation AI model “Gemini Omni.”
- This latest video generation AI model supports up to 15-second video generation and prompt-based interactive multi-turn editing features, while significantly improving character and object consistency (reference adherence).
- It is now generally available in “Google AI Studio” and the “Gemini API,” and will also be rolled out sequentially for “Google AI Ultra” members in the “Google Cloud” console, the AI video editing service “Google Vids,” the AI video production tool “Google Flow,” and the “Gemini” app.
On Thursday, November 5, 2026 (local time), Google officially announced “Gemini Omni 1.2 Flash,” the latest version of its video generation AI model “Gemini Omni.”
“Gemini Omni 1.2 Flash” is the latest video generation AI model that supports up to 15-second video generation and prompt-based interactive multi-turn editing features, along with significantly improved character and object consistency (reference adherence). Furthermore, its price has been reduced by 30% compared to the previous-generation “Gemini Omni 1.1 Flash”, becoming available at $0.07 per second for 720p video.
“Gemini Omni 1.2 Flash” is now generally available on the developer platforms “Google AI Studio” and the “Gemini API,” and will also be rolled out sequentially in the “Google Cloud” console, the AI video editing service “Google Vids,” the AI video production tool “Google Flow,” and the “Gemini” app for “Google AI Ultra” members.
Up to 15-Second Video Generation
“Gemini Omni 1.2 Flash” expands the previous maximum video generation length of up to 10 seconds, natively supporting video generation and editing of up to 15 seconds.
Video length can be specified in fine 1-second increments between 3 and 15 seconds, making it possible to complete short social media ads or product explainer videos in a single generation.
Interactive Multi-Turn Editing
“Gemini Omni 1.2 Flash” supports multi-turn editing, allowing users to give successive conversational instructions to previously generated videos and make incremental modifications.
Without having to recreate a video from scratch, you can instruct changes to the background, adjustments to lighting or weather, or switches in aspect ratio while maintaining character actions and camera angles. This allows you to quickly produce varied videos for different target audiences or ad slots from the same base video.
Enhanced Reference Adherence and Consistency
“Gemini Omni 1.2 Flash” significantly enhances consistency when combining multiple reference inputs, such as characters, products, style images, motion clips of up to 3 seconds, and audio.
Even when camera angles shift, facial features, costume drapes, packaging typography, and brand colors are faithfully preserved from the first frame to the last without flickering. It is also ideal for generating virtual try-on videos and catalog videos containing 360-degree turns from still product photographs.
More Natural Conversational Audio and Cinematic Camera Work
“Gemini Omni 1.2 Flash” has improved not only visuals, but also the expressive power of audio and camera work.
In scenes featuring multiple characters, it achieves more fluent, natural speech and high-precision lip-syncing. Furthermore, it has evolved to smoothly depict complex camera movements like FPV drone flights through buildings, dynamic zoom-outs, and panning in a cinematic manner while suppressing unnatural jump cuts.
Source: Google (https://blog.google/innovation-and-ai/technology/developers-tools/introducing-gemini-omni-1-2-flash/)