Gemini Omni 1.2 Flash launched with 15-second video generation and AI editing

Add MobiGyaan as a preferred source on Google

Compared to Gemini Omni 1.1, the new version supports native video generation and editing for clips of up to 15 seconds. Google has also introduced improvements to reference adherence, speech generation, and camera controls, giving users more flexibility when creating and refining videos.

What’s New In Gemini Omni 1.2 Flash?

Longer Videos & Improved Consistency

Gemini Omni 1.2 Flash can generate and edit videos ranging from 3 to 15 seconds, with duration controls available in one-second increments.

The model is designed to maintain character, object and scene consistency across frames, reducing visual drift and morphing during movement and camera transitions. This consistency also applies when interpolating between the first and last keyframes or extending an existing clip into a longer sequence.

This can help preserve the appearance of products, characters, props and environments throughout a video.

Improved Reference Image & Motion Adherence

The updated model can combine character, product and style reference images with short motion clips of up to three seconds and audio cues.

It is designed to retain visual details such as faces, garments, packaging typography and brand colors across different camera angles. Google highlights potential applications including virtual try-on clips, product catalog spins and branded video advertisements created from existing product photographs or character sheets.

Multi-Turn Conversational Video Editing

One of the key additions is multi-turn conversational editing. Users can refine an existing video through multiple prompts instead of generating a new clip from scratch.

Editing options include:

• Replacing products or backgrounds.
• Adjusting lighting and weather conditions.
• Changing the video aspect ratio.

These edits work across full 15-second clips, allowing users to create different versions of an advertisement for different audiences and placements without resubmitting the original instructions.

Improved Speech Generation & Camera Controls

Gemini Omni 1.2 Flash also brings more fluid dialogue, clearer lip-sync and distinct multi-speaker conditioning.

Google highlights several applications for these improvements, including multi-character dialogue in Flow, video explainers in Gemini for business, and videos with avatars, music and narration in Google Vids.

The model also supports continuous camera movements such as sweeping FPV flythroughs, dramatic zoom-outs and camera orbits. Google says these improvements help reduce unwanted jump cuts and maintain steadier scale during camera movements.

Google Vids Gets Gemini Omni 1.2 Flash

Google Vids, the company’s video creation and editing tool, is also receiving the new model for video generation.

Starting today, anyone with a Google account or Google Workspace subscription can generate up to three 15-second videos per month at no cost in Google Vids.

The integration supports three main use cases:

• Product promotions: Create social media content and promotional materials in a single generation without cuts.
• Image-to-video generation: Upload reference images from trips, school projects, or learning materials and turn them into videos while maintaining character and scene consistency.
• Speech narration: Generate clips with more natural-sounding speech for social media content, explainers and teaser videos.

Gemini Omni 1.2 Flash Pricing & Availability

Google has reduced the price of 720p video generation by 30% compared to Gemini Omni 1.1. The new pricing is $0.07 per second.

Users can also manage generation costs by creating initial drafts at 360p or 720p, making changes through follow-up edits instead of generating an entire clip again, and upscaling selected videos to 1080p or 4K.

Gemini Omni 1.2 Flash is generally available through the Gemini API and Gemini for business. Developers can prototype with the model through Google AI Studio, while businesses can access it through the Google Cloud console and build enterprise solutions using the API.

Users can also access the model across Google’s applications:

• Developers: Use the Gemini API in Google AI Studio, explore interactive starter apps, and refer to the developer documentation and pricing.
• Businesses: Access the model through Google Cloud and use its API for enterprise video solutions.
• Flow: Direct scenes and make multi-turn edits using the updated model.
• Google Vids: Generate up to three 15-second videos per month at no cost.
• Gemini app: Create 15-second videos, with rollout underway for Google AI Ultra subscribers.
• Google Ads: Businesses can use the model to scale video campaigns.

The release is supported by Google Cloud’s enterprise-grade infrastructure and security. Provisioned Throughput is also available, while generated outputs include an imperceptible SynthID watermark.

Google says it is continuing to improve the model in areas such as fast sports action, rapid script changes, and complex physics.