Google Launches Gemini Omni Flash Video Generation AI as Public API

Google has begun providing Gemini Omni Flash, a video generation AI model, as an API to developers and enterprises. The model was first introduced to general users at Google I/O 2026, and the API release now enables integration into corporate business systems. Gemini Omni Flash is positioned as the first model in Google's newly launched Omni family.

Previously, enterprises required multiple steps to produce short internal videos. The entire workflow—from scripting to filming to editing to revision—was time-consuming and costly, and even a single-line text modification mandated by legal review required redoing all stages. Due to these burdens, many internal videos were never produced at all. Even after generative AI became widespread, the typical approach was to combine multiple tools: script generation, image generation, video conversion, text-to-speech, and lip-sync software. This created separate contractual relationships and data management requirements for each tool.

Gemini Omni Flash is designed to consolidate the roles of these multiple tools into a single model. It accepts text, images, and video as input and outputs a completed video with synchronized audio. Notably, it features a conversational editing capability that allows users to iteratively refine completed videos through chat-like instructions. Changes to lighting, frame adjustments, and costume replacements can be executed without starting from scratch.

Leveraging reference images is another key feature of the model. In addition to text instructions, users can provide multiple reference images or existing video clips as input. When given a specific product photo, the model generates video that replicates its color palette and shape. While pixel-perfect accuracy is not guaranteed, it maintains a level of precision sufficient for identification. This capability opens possibilities for using product photos and brand logos as materials in video production.

Before the API release, Omni Flash was limited to Google's consumer services and professional/amateur tools. There was no means to access it programmatically, creating constraints for teams within organizations that produce the most videos—such as marketing and learning & development (L&D) departments. The API offering now enables these departments to integrate the model into their own workflows and systems.

The significance of consolidating multiple tools into one extends beyond technical considerations. It generates operational benefits including reduced vendor numbers, unified data management rules, and simplified compliance procedures. For enterprises cautious about generative AI adoption, the lower cost of tool consolidation can reduce the barrier to considering implementation. Scenarios in which in-house video production becomes a realistic option may become increasingly common.

Gemini Omni Flash chose video as the starting point for the Omni family's goal of generating everything from any input. How this family will expand to other media formats and enterprise use cases is a key point of interest. Whether the conversational editing interface becomes established in corporate video production workflows remains to be seen through actual operational experience.