AI News (2026/11/6): Google Launches Gemini Omni 1.2 Flash Video Generation Model and Opens API to Promote Video Generation Technology Adoption
Executive Summary:
Google announced the release of Gemini Omni 1.2 Flash video generation model on November 6, simultaneously opening its API interface. This model continues the multimodal processing advantages of the Gemini series while achieving breakthrough optimizations in video generation speed and cost control.
Key Details
Google released Gemini Omni 1.2 Flash video generation model on November 6 alongside its API interface. The model maintains the multimodal processing strengths of the Gemini series while delivering significant improvements in video generation speed and cost efficiency.
- [Multi-turn Dialog Editing]: Supports iterative modifications through natural language instructions (e.g., adjusting camera angles or adding effects). The system can parse sequential contextual commands while preserving previous editing states.
- [Strong Consistency in First and Last Frames]: Utilizes enhanced temporal-spatial alignment algorithm (Temporal-Spatial Alignment v3) to ensure semantic coherence and visual style consistency between video opening and closing frames.
- [Cost Structure Optimization]: Generation cost for 720p resolution videos reduced to $0.07 per second (30% decrease from version 1.1). Supports pay-as-you-go pricing model and batch processing discounts.
AI-ALL In-depth Analysis
The launch of Gemini Omni 1.2 Flash signifies AI video generation's transition into industrial production. Its multi-turn dialog editing capability will significantly reduce interaction costs, creating differentiated advantages in advertising production and game development scenarios. The implementation of strong first-last frame consistency addresses semantic drift issues in long-sequence generation, establishing a reliable technical foundation for post-production in film and television.
Notably, this model enhances audio guidance processing through coupled training of speech fingerprint feature extraction and lip movement modeling (Speech-Visual Coupling Module), further lowering barriers for virtual anchor and digital human development. The $0.07 per second pricing strategy could disrupt existing video creation toolchains (e.g., Runway Gen-3 / Pika v4), particularly establishing cost advantages in short-form video content production.
The open-source community has already begun examining its modular API architecture design (Modular API Architecture). Developers can construct complex workflows (Agentic Workflow) by combining text/image/audio modules. However, limitations in cloud deployment model coverage for localized inference scenarios should be noted - current version only supports Google Cloud Platform deployment architecture.
