Gemini Omni 1.2 Flash

Today, we’re rolling out Gemini Omni 1.2 Flash, and making it mostly accessible throughout the Gemini API and Gemini for business. Gemini Omni 1.2 Flash introduces native 15-second video creation, multi-turn conversational editing capabilities, and enhanced citation following. We have dropped the cost 30% from Gemini Omni 1.1 making it accessible at $0.07 per second for 720p video.

Here’s an overview of what’s new:

Fast and cost-effective at scale

Beyond the lesser price, the workflow keeps expenses down: you can outline accelerated at 360p or 720p, fix a item alongside a follow-up edit alternatively of re-generating the complete clip, and afterward upscale your finest creations to 1080p or 4K.

Longer clips up to 15 seconds

You can now create and edit videos natively up to 15 seconds long, up from 10 seconds alongside Omni 1.1. You can additionally define duration in one-second increments from 3 to 15 seconds, giving you area to inform complete stories, create merchandise explainers, or create complete social ad spots in a sole generation.

Better citation adherence

You can guide generation combining character, product, or manner images, abbreviated motion clips (up to three seconds), and audio cues — and Omni 1.2 Flash holds onto those particulars from commencement to finish. Faces, garments, packaging typography, and brand colors remain accordant throughout camera angles, making it uncomplicated to rotate existing merchandise photography or character sheets into virtual try-on clips, catalog spins, and tagged video ads.

Combining one example photo alongside front-and-back merchandise shots of three outfits, Gemini Omni 1.2 Flash generates a 15-second virtual try-on sequence throughout 360-degree turns and equivalent cuts during keeping the model’s character and all garment’s cut, drape, and rear particulars consistent.

More accordant results

You can keep stronger character, object, and environment consistency from the archetypal to the final frame, alongside small ocular drift or morphing throughout motion and camera transitions. Whether you’re interpolating between a archetypal and final keyframe, or extending a clip to build a longer sequence, products, characters, props, and environments remain accordant from attempt to shot.

Given the identical fixed studio lighting showcase poster and five-layer animation prompt, Gemini Omni 1.1 Flash does not equivalent the citation depiction as closely as 1.2 does. Gemini Omni 1.2 Flash builds the poster sequentially tier by tier on the dusty-pink cyclorama rising the hexagonal plinth, assembling the ribbed glass, heated filament, faceted crystal, and coiled-cord letterforms for HALO, GLOW, and LUMEN, and locking all material, shadow, and typographic item consistently through the final frame.

Easier editing in conversation

You can refine the identical clip rotate by rotate — specified as swapping a merchandise or background, changing the lighting or weather, or switching the aspect proportion — alternatively of starting complete from scratch. Multi-turn conversational editing plant throughout complete 15-second clips, making it uncomplicated to create ad variants for distinct audiences and placements without re-briefing or re-generating.

Starting from one generated clip, multi-turn conversational prompts toggle the backdrop rotate by rotate — from a striped refuge shelter to an oak forest, a stable alongside a pale horse, a market fair, and a rock castle — during keeping the two characters’ performances, armor reflections, and camera framing intact.

More natural speech

You can create scenes alongside additional fluid, natural-sounding dialogue, clearer lip-sync, and distinct multi-speaker conditioning. This volition display up whether you’re directing multi-character conversation in Flow, creating video explainers in Gemini for business, or creating polished, pro videos alongside avatars, music and narration in Google Vids.

Conditioned on one conservatory environment depiction alongside four character, mouth, and one-second sound references, Gemini Omni 1.2 Flash routes all sampled sound to the correct character alongside expressive lip-sync and natural comedic timing throughout a 15-second animated scene.

More controllable, cinematic look

You can straightforward complex, uninterrupted camera movements akin sweeping FPV flythroughs to theatrical zoom-outs and orbits alongside considerably small unwanted jump cuts, steadier scale, and a additional polished, cinematic look.

Guided by two architectural citation photos, Gemini Omni 1.2 Flash directs one uninterrupted FPV drone escape throughout the piazza, through the bronze doors, and up through the Pantheon’s coffered dome to the eye alongside zero jump cuts.

Longer clips alongside smoother motion and tighter consistency

Gemini Omni 1.2 Flash generates clips up to 15 seconds lengthy that clasp references together improved than before. You can scroll through to see characters, text, and camera moves remain accordant throughout longer shots, multiple turns and transitions alongside no cuts.

Get started alongside Gemini Omni 1.2 Flash

While betterment is ongoing to additional enhance things akin accelerated sports action, fast manuscript changes and complex physics, Gemini Omni 1.2 Flash is ideal for businesses, developers, and creators producing videos at scale. You can get started alongside Gemini Omni 1.2 Flash today as it rolls out throughout Google’s products and developer platforms:

• For developers using the Gemini API: Prototype in Google AI Studio, examine interactive starter apps, and inspect out the developer documentation and pricing.
• For businesses: Try it in the Google Cloud console, and build endeavor solutions alongside the API.
• For everyone using Google apps: Direct scenes and multi-turn edits in Flow, create 15-second clips at no additional disbursal in Google Vids alongside a Google document or Google Workspace subscription, create 15-second videos in the Gemini app (rolling out to Google AI Ultra subscribers), and measure video campaigns in Google Ads.