Google Releases Gemini Omni 1.2 Flash: 15-Second AI Video at $0.07 per Second
By Asif Razzaq — November 5, 2026
Google has released Gemini Omni 1.2 Flash, the newest version of its video generation and editing model. Gemini Omni 1.2 Flash generates and edits clips up to 15 seconds natively and cuts the price by 30% compared with Omni 1.1 Flash. The model is now generally available on Google Cloud and the Gemini API. For developers, the key takeaways are longer clips, fewer re-renders, and a lower cost per second.
TL;DR
- Runs on: Google-hosted infrastructure only (Gemini API, Google Cloud). No open weights, no local deployment.
- Performance: Google reports better reference adherence and less drift. No benchmark scores for version 1.2 have been published.
- Best for: Native 15-second clips at $0.07 per second of 720p video — about $1.05 per 15-second clip.
- Limitations: Google acknowledges that fast sports action, rapid script changes, and complex physics still need work.
What Is Gemini Omni 1.2 Flash?
Gemini Omni 1.2 Flash is a multimodal video model from Google DeepMind. It accepts text, images, short video clips, and audio as input, and it outputs video with audio. Google's developer documentation describes the Omni line as natively multimodal, with editing handled through the Interactions API.
The release lands amid intensifying competition in the AI video space, where 2026 has seen a wave of models pushing toward longer native clip lengths and lower per-second pricing. Google's positioning of Omni 1.2 Flash — longer output combined with a 30% cost reduction — targets production workflows where re-rendering for length has historically inflated costs.
Key Capabilities
- Native 15-second generation: Clips are generated at full length without stitching together shorter segments, reducing re-renders.
- Built-in editing: The model handles video edits natively through the Interactions API.
- Multimodal input: Accepts text, images, short video clips, and audio, producing video with synchronized audio output.
- Managed deployment: Available exclusively through Google-hosted services (Gemini API and Google Cloud), with no open-weight or on-premises options.
Performance and Limitations
Google reports improved reference adherence and reduced drift compared with earlier versions, though the company has not published benchmark scores for version 1.2. Google also notes that the model still struggles with fast sports action, rapid script changes, and complex physics simulations.
Originally reported by Marktechpost.
via MarkTechPost