Google’s Gemini Omni 1.2 Flash Brings 15-Second AI Video Generation to Its API

November 11, 2026

Google has made Gemini Omni 1.2 Flash generally available through the Gemini API and Gemini for business, adding longer AI-generated video clips, conversational editing and stronger reference consistency. The release is significant because it targets the production workflow—not just one-off demonstrations—with lower pricing and iterative edits designed to reduce the cost of making variations.

What Google changed

Gemini Omni 1.2 Flash can generate and edit video clips up to 15 seconds, compared with 10 seconds for Omni 1.1. Google says users can specify durations from three to 15 seconds in one-second increments. The model is priced at $0.07 per second for 720p video, which Google describes as a 30% reduction from Omni 1.1.

The model also supports multi-turn editing. A creator can change a background, lighting setup or aspect ratio in a follow-up instruction instead of regenerating the entire clip from scratch. Google says the workflow can begin with faster 360p drafts, move to 720p, and upscale selected work to 1080p or 4K.

Why reference consistency matters

Generative video has often struggled to keep people, products and environments stable through camera movement. Google says Omni 1.2 Flash can combine reference images, short motion clips and audio cues while preserving details such as faces, clothing, packaging typography and brand colours across a sequence.

That claim should be read as a product capability, not a guarantee that every prompt will produce production-ready footage. Google lists fast sports action, rapid script changes and complex physics among areas that still need improvement.

Who may benefit

The immediate audience includes developers building video workflows, agencies producing multiple ad variants, and businesses that need short explainers or product demonstrations. The lower per-second price and follow-up editing model could matter more to these users than a small increase in maximum clip length.

For enterprises, Google says the model is backed by Google Cloud infrastructure and security, with Provisioned Throughput available. Outputs carry an imperceptible SynthID watermark. Those controls address deployment and provenance concerns, but they do not remove the need for human review, rights clearance or disclosure when synthetic media could be mistaken for documentary footage.

The broader shift

Google’s release shows the market moving toward controllable, iterative video systems. The competitive question is no longer only whether a model can create a plausible clip. It is whether users can maintain continuity, revise a result without starting over, and produce enough variants at a predictable cost to justify using the tool in a real workflow.

That makes Omni 1.2 Flash relevant to both creators and software teams. Its long-term value will depend on reliability in ordinary production conditions, not on the most polished examples in a launch announcement.

Sources

Topics:
Read next
November 20, 2026

U.S. Expands AI-Chip Controls to Remote Cloud Access as China Pushes Domestic Alternatives

A new U.S. interim rule extends advanced AI-computing controls to remote access through overseas data centres, raising compliance and infrastructure questions for cloud providers and Chinese technology companies.
November 20, 2026

OpenAI Releases GPT-6.2 Astra and Sol as AI Competition Shifts to Workflow Economics

OpenAI has released GPT-6.2 Astra and Sol with lower API pricing and stronger professional-work claims. Early independent reactions are mixed, making task-level testing more important than benchmark headlines.