Google Launches Gemini Omni 1.2 Flash for 15-Second AI Video Generation
Google has released Gemini Omni 1.2 Flash, an AI video model capable of generating 15-second clips, a 50% increase from its predecessor. Priced at $0.07 per second, it offers improved visual consistency and multi-turn editing. The model is now generally available via Google Cloud and consumer apps, marking a significant step in accessible AI video production.
Key points
- Google DeepMind released Gemini Omni 1.2 Flash, its latest AI video generation model, which supports clips up to 15 seconds long.
- The new model costs approximately $0.07 per second for 720p video, representing a 30% price reduction from the previous $0.10 rate.
- Users can edit video content through multi-turn conversational commands, allowing changes to lighting, backgrounds, or objects without regenerating the entire clip.
- Internal testing indicates a 62% preference for 1.2 over version 1.1 due to better visual consistency in faces, clothing, and brand elements.
- The model is generally available through Google Cloud for businesses, while consumer access is provided via the Gemini app, Flow, and Google Vids.
Google Expands AI Video Capabilities with New Model
Google has officially made its latest artificial intelligence video generation model, Gemini Omni 1.2 Flash, generally available. This release marks a significant update to the company's video generation suite, introducing longer clip durations, reduced costs, and enhanced editing capabilities. The model builds upon the foundation laid by Gemini Omni 1.1 Flash, which was initially showcased at Google's I/O conference in May, by addressing previous limitations in duration and user control.
Pricing and Accessibility Changes
A key feature of the 1.2 release is a substantial reduction in pricing. Google has lowered the cost of generating 720p video to approximately $0.07 per second. This represents a 30% decrease from the $0.10 per second rate established in June. The model allows users to specify video lengths in one-second increments, ranging from three to 15 seconds. This flexibility provides creators with more granular control over output duration compared to the fixed limits of earlier versions.
Technical Improvements and Consistency
The new model significantly extends the maximum video length from 10 seconds to 15 seconds. Google states that this increase allows for more complex scenes and narratives. Beyond duration, the model promises improved visual consistency. According to Google DeepMind, the system now maintains better adherence to reference images, keeping faces, clothing, packaging text, and brand colors consistent across different camera angles. Internal comparisons conducted by Google testers showed a 62% preference for version 1.2 over version 1.1, citing fewer glitches and jump cuts.
Editing and User Interaction
Gemini Omni 1.2 Flash introduces multi-turn conversational editing. Users can now refine generated videos through iterative chat interactions rather than regenerating clips from scratch. This allows for specific adjustments, such as swapping a product in the frame, altering the background, or modifying lighting conditions. This feature aims to streamline the creative workflow, reducing the time and computational resources required to achieve the desired visual outcome.
Platform Availability and Distribution
The model is being rolled out across multiple Google platforms. For businesses, it is available through Google Cloud, marking the first time the Omni series has been generally accessible via this channel. Consumer users can access the model through the Gemini app, specifically those with Google AI Ultra subscriptions. Additionally, free users can experiment with some capabilities through Google Flow and Google Vids. The model is also integrated into Google Ads, potentially impacting digital advertising production.
Industry Context and Future Development
The release places Google in direct competition with other major tech firms advancing AI video generation. For instance, ByteDance’s Seedance 2.5, unveiled in June, supports 30-second clips. Google acknowledges that the model still faces challenges with fast sports action, quick script changes, and complex physics simulations. The company indicates that work is ongoing to address these limitations in future updates.
Key takeaways
- Google released Gemini Omni 1.2 Flash, enabling 15-second AI video generation.
- The price per second for 720p video dropped by 30% to $0.07.
- Multi-turn editing allows users to modify videos via chat commands.
- Internal tests show a 62% preference for the new model's visual consistency.
- The model is available via Google Cloud, Gemini app, Flow, and Google Vids.
- Google continues to work on improving physics and fast-action rendering.
Sources
The WireByte editorial team synthesises technology news from multiple primary sources, verifies the facts, and links every source. Articles are produced with AI assistance and reviewed under our editorial policy.
