Quick Answer
Gemini Omni 1.2 Flash is Google’s new AI video model for making and editing clips up to 15 seconds, with 720p generation priced at $0.07 per second. Google says the model also supports follow-up edits and stronger reference consistency. Use lower-resolution drafts first, then upscale approved clips, while treating person-speech edits and synthetic-media disclosure as material safeguards.
Key Takeaways
- Gemini Omni 1.2 Flash generates and edits clips from 3 to 15 seconds long.
- Google cut 720p generation pricing by 30% to $0.07 per second.
- Multi-turn editing lets users revise an existing clip with follow-up instructions.
- Google says the model better preserves visual and audio references across shots.
- Generated output includes an imperceptible SynthID watermark, according to Google.
What is Gemini Omni 1.2 Flash?
Gemini Omni 1.2 Flash is Google’s generally available AI video model for generating and editing short clips. Google DeepMind announced the model on November 5, 2026, and said it is available through the Gemini API and Gemini for business. Google’s Gemini Omni 1.2 Flash announcement identifies video generation, editing, and reference handling as the central changes.
Gemini Omni 1.2 Flash can natively create and edit clips from 3 to 15 seconds long. The 15-second limit extends the previous Omni 1.1 Flash limit of 10 seconds, which gives users more room for a complete product demonstration, scene transition, or short explanatory sequence without splitting the idea into separate generations.
Gemini Omni 1.2 Flash is still designed for short-form work rather than a replacement for conventional long-form video editing. The practical advantage is a longer single generation, not an unlimited timeline. Teams planning a longer sequence should break the work into deliberate shots, review each result, and assemble approved clips in an editing workflow.
How does Gemini Omni 1.2 Flash change video editing?
Gemini Omni 1.2 Flash adds multi-turn conversational editing, which lets users revise a clip through follow-up instructions instead of starting a new generation from scratch. Google says a user can make a request, examine the result, and then direct the model to change the existing clip. That workflow can preserve useful elements while addressing a specific problem in a later prompt.
Gemini Omni 1.2 Flash also aims to hold image, video, and audio references more consistently across shots. Google says the improved reference behavior includes faces, garments, product packaging, and brand colors. This matters for work that needs a recurring visual subject, although Google’s statement is a product claim and does not guarantee that every generated shot will preserve every detail exactly.
Creators working with product imagery should review every frame before publishing because a reference feature is not the same as factual verification. A package color, garment detail, or visual likeness can matter in advertising and customer-facing material. Teams that use editable AI graphics alongside generated footage should also keep approved source assets separate from experimental outputs.
Why does the 15-second generation limit matter?
Gemini Omni 1.2 Flash supports clips up to 15 seconds, which is 5 seconds longer than the prior Omni 1.1 Flash limit. The added duration matters because a single short clip can now contain more complete action, including an opening movement, a product interaction, and a closing frame, without requiring an immediate cut between separately generated segments.
Gemini Omni 1.2 Flash does not make shot planning unnecessary. Longer clips can create more continuity to review, particularly when a video includes a recognizable person, a branded object, or synchronized sound. The practical response is to define the intended beginning, middle, and end before generating the clip, then check whether the visual and audio elements remain appropriate throughout the full duration.
| Gemini video capability | Omni 1.1 Flash | Omni 1.2 Flash | What changes for users |
|---|---|---|---|
| Native clip length | Up to 10 seconds | 3 to 15 seconds | Users can generate a longer single shot. |
| Editing workflow | Not listed in the supplied launch details | Multi-turn conversational editing | Users can request revisions to an existing clip. |
| 720p generation price | Higher than the new price | $0.07 per second | Google says the new rate is 30% lower. |
| Output options | Not listed in the supplied launch details | Draft at 360p or 720p, upscale to 1080p or 4K | Users can reserve higher resolution for selected clips. |
Gemini Omni 1.2 Flash therefore provides a longer unit of generation, not a guarantee of a finished video. The most sensible approach is to use the 15-second ceiling when a scene benefits from continuity and use shorter clips when a project needs precise control over each shot.
What does Gemini Omni 1.2 Flash cost to use?
Gemini Omni 1.2 Flash costs $0.07 per second for 720p video generation, according to Google. Google says that price is 30% lower than Omni 1.1 Flash’s 720p generation rate. At the stated rate, a 15-second 720p clip equals $1.05 before considering any other services, features, or account terms that may apply.
Gemini Omni 1.2 Flash also lets users draft clips at 360p or 720p and upscale selected clips to 1080p or 4K. This matters because lower-resolution drafts can help users decide whether the composition and motion are worth keeping before committing to a higher-resolution version. Google’s announcement does not establish a single price in the supplied information for every draft and upscale option, so users should confirm the applicable pricing before planning a high-volume workflow.
Gemini Omni 1.2 Flash’s lower 720p rate is most useful for iterative work, where users may need several variations before selecting one. The practical cost control is to write a specific prompt, create a lower-resolution draft, make conversational edits to the promising result, and upscale only the clip approved for final use.
Where is Gemini Omni 1.2 Flash available?
Gemini Omni 1.2 Flash is generally available through the Gemini API and Gemini for business, according to Google DeepMind. Google also says the model is rolling out in Flow, Google Vids, Google Ads, and the Gemini app. Gemini app access is rolling out to Google AI Ultra subscribers, which means availability can differ by product and account type during the rollout.
Google Vids availability is rolling out over 1 to 3 days for Rapid Release and Scheduled Release domains. The rollout includes personal Google accounts, Google AI Pro and Ultra subscribers, and listed Google Workspace plans, according to Google Workspace’s Vids update. The short rollout window is useful context, but users should still check the relevant service because a feature can appear at different times across accounts.
Gemini Omni 1.2 Flash access also depends on where a user works. A business using the API may have a different path from an individual using the Gemini app or a team creating presentations in Google Vids. Organizations configuring Gemini tools should review Gemini app permissions before connecting data or services to an AI workflow.
What are the safety and disclosure limits for Gemini Omni 1.2 Flash?
Gemini Omni 1.2 Flash can change a person’s speech while editing video, which creates a meaningful synthetic-media risk when the output depicts a real person. Google’s model card documents that capability, and Google says generated outputs carry an imperceptible SynthID watermark. Google’s Gemini Omni 1.2 Flash model card provides the stated capability and watermark information.
Gemini Omni 1.2 Flash’s watermark is a provenance measure, not a substitute for responsible use or clear disclosure. An imperceptible marker may help identify AI-generated material through compatible detection methods, but viewers cannot rely on seeing a visible label in ordinary playback. The practical response is to disclose materially altered or AI-generated footage clearly, especially when it includes recognizable people, product claims, or realistic speech.
Gemini Omni 1.2 Flash also has an audio-reference limitation. Google says audio-reference use is allowlist-gated, so access to that part of the model is not described as open to every user. Users should not assume that a feature shown in an announcement will appear in every account, and they should obtain appropriate permission before submitting any person’s likeness, voice, or proprietary material.
How should creators use Gemini Omni 1.2 Flash efficiently?
Gemini Omni 1.2 Flash works best when creators treat generation as a reviewable production step rather than an automatic publishing tool. The model’s 360p and 720p drafting options, multi-turn editing, and selective upscaling support a staged process. That structure can limit unnecessary higher-resolution work while giving the creator a chance to correct the clip before finalizing it.
- Define the shot’s subject, action, setting, and intended duration before generating it.
- Generate an initial draft at 360p or 720p.
- Review the clip for visual consistency, speech changes, packaging details, and brand-color accuracy.
- Request a specific follow-up edit when the initial result is close but needs revision.
- Upscale the approved clip to 1080p or 4K when the final project requires it.
- Disclose AI-generated or materially altered footage before publication when viewers could reasonably mistake it for unaltered media.
Gemini Omni 1.2 Flash does not remove the need for human review because generated video can contain meaningful visual or speech changes. Stop before publishing if a clip changes a real person’s words, makes a product claim that has not been verified, or uses material without permission. A legal, brand, or communications reviewer should approve high-stakes advertising, political, financial, medical, or public-facing content.
FAQ
What is the maximum Gemini Omni 1.2 Flash video length?
Gemini Omni 1.2 Flash can natively generate and edit clips from 3 to 15 seconds long. Google says the prior Omni 1.1 Flash model had a 10-second generation limit.
How much does Gemini Omni 1.2 Flash cost at 720p?
Gemini Omni 1.2 Flash costs $0.07 per second for 720p generation, according to Google. A 15-second clip at that stated rate equals $1.05 before any other applicable service terms or charges.
Can Gemini Omni 1.2 Flash edit an existing clip?
Gemini Omni 1.2 Flash supports multi-turn conversational editing, so users can request follow-up changes to a clip. Google says the feature is intended to let users revise a result instead of regenerating it from scratch.
Can Gemini Omni 1.2 Flash generate 4K video?
Gemini Omni 1.2 Flash lets users draft clips at 360p or 720p and upscale selected clips to 1080p or 4K. Users should confirm applicable pricing and availability before relying on an upscale option for production work.
Does Gemini Omni 1.2 Flash watermark generated video?
Gemini Omni 1.2 Flash outputs carry an imperceptible SynthID watermark, according to Google. The watermark supports provenance efforts, but creators should still disclose realistic AI-generated or materially altered video when the context requires it.