🔍 Read the full analysis: Longer Clips, Lower Costs: What Gemini Omni 1.2 Flash Offers on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Google announced Gemini Omni 1.2 Flash on Nov. 5, offering video generation and editing for clips up to 15 seconds. Google lists 720p video at $0.07 per second, 30% below the stated price for version 1.1; independent tests and prices for other resolutions were not provided.
Google announced Gemini Omni 1.2 Flash on Nov. 5, an AI video model that generates and edits clips up to 15 seconds long. The company says 720p video costs $0.07 per second, 30% less than version 1.1, and that the model is generally available through the Gemini API and Gemini for business.
The new model raises the maximum native clip length from 10 seconds in version 1.1 to 15 seconds. Google says users can choose a duration in one-second increments from three to 15 seconds. The company also describes multi-turn editing: users can follow up with prompts to adjust a clip, such as changing its background or lighting, instead of generating the entire video again.
Google says the model can draw on reference images, motion clips up to three seconds, and audio cues. It describes improvements in following those references, keeping characters, products and scenes consistent across frames, and generating speech, lip-sync and camera movement. These are claims in Google’s product announcement, which did not include independent tests or standardized comparisons.
The stated $0.07-per-second rate applies to 720p. Google says users can draft at 360p or 720p and upscale selected clips to 1080p or 4K, but the supplied announcement does not state prices for those resolutions or explain how upscaling is billed. Google also says outputs carry an imperceptible SynthID watermark and that Provisioned Throughput is available.
Gemini Omni 1.2 Flash Longer clips, lower stated cost
Google’s update brings longer video generation and conversational editing. Its price and quality claims still need independent evaluation.
What changes for creators
The update targets short video workflows that need multiple versions and revisions.
More room for a scene
The maximum rises from 10 to 15 seconds, with duration selectable in one-second increments.
Revise with prompts
Users can request changes such as a new background or lighting adjustment without regenerating the whole clip.
Guide the generation
Google says the model can use reference images, motion clips up to three seconds, and audio cues.
Where it is rolling out
Google says access is beginning across its products and developer platforms.
What the price covers
The announced rate applies to 720p video; other resolution and upscaling costs were not provided.
Google says creators can choose either resolution for drafts.
The stated $0.07-per-second rate is for 720p video.
Selected clips can be upscaled to 1080p or 4K, but those prices were not stated.
Claims awaiting testing
Google describes improvements, but the announcement supplies no independent benchmarks or standardized comparisons.
- Better reference following and continuity
- Improvements to speech, lip-sync, and camera movement
- Edits can preserve scene details across changes
- Reliability across varied scenes and demanding footage
- Failure rates and time saved in real workflows
- Total cost, including higher resolutions and upscaling
A Lower Rate for Short Video
The update combines a longer clip limit, editing through follow-up prompts and a lower stated 720p rate. For developers, businesses and creators producing short ads, explainers or social videos, those features could make it easier to make variations without rebuilding each clip. If edits preserve the scene and details as Google says, multi-turn revisions may reduce repeated generation.
The practical value, however, depends on output quality and the full cost of a workflow. Google did not quantify how often edits retain requested details, how much time they save, or the cost of higher-resolution output. The 30% reduction compares version 1.2 with version 1.1; it is not a comparison with other video services.
Products Worth Considering
Apple MacBook Air 13.6-inch with M5 Chip, 16GB Memory, 512GB SSD, Sky Blue
Apple MacBook Air 15-inch with M5 chip, 24GB RAM, 512GB SSD, Midnight
Samsung Galaxy Tab A11 Wi‑Fi (International Version), 8.7, 64GB + 4GB, Gray
As an affiliate, we earn on qualifying purchases.
How Version 1.2 Changes
Gemini Omni 1.2 Flash follows Gemini Omni 1.1, which Google says supported clips of up to 10 seconds. The new 15-second ceiling adds five seconds to the maximum native clip length, while multi-turn editing lets users request changes to a generated result. Google presents these changes alongside claimed gains in reference adherence, continuity and speech.
The company cited possible uses such as virtual try-on sequences, animated product posters and scenes with multiple speaking characters. Those examples show intended applications, not independent evaluations of how well the model handles them. Google also says it is still working on fast sports action, rapid script changes and complex physics, indicating areas where performance remains under development.
“Gemini Omni 1.2 Flash introduces native 15-second video creation, multi-turn conversational editing capabilities, and enhanced reference following.”
— Google DeepMind, in its Nov. 5 product announcement
Pricing and Performance Gaps
There are no independent benchmarks or detailed testing methods in the announcement to establish how much reference following, continuity, speech or lip-sync improved over version 1.1. Google did not report failure rates, or explain how reliably edits preserve faces, text, product details or scene elements in demanding projects. Its acknowledgment that work continues on fast action, rapid script changes and complex physics points to known areas still being developed.
Practical access and total cost are also not fully specified. The announcement gives no prices for 360p, 1080p or 4K, and no billing details for upscaling. Google says the model is rolling out across its platforms but does not provide a schedule by product, region or account. The extent of the 30% price difference is clear only against the stated version 1.1 rate, not against competing services.
Rollout and Independent Testing
Google says Gemini Omni 1.2 Flash is rolling out through the Gemini API and Gemini for business, and points developers to Google AI Studio for prototyping. The announcement also names Flow and Google Vids among the products where the model is available, but does not give a timetable for access across every platform or account.
The next useful evidence will be independent testing on varied footage, including difficult motion, detailed products and scenes with multiple characters. Clearer pricing for resolutions beyond 720p and for upscaling will help users calculate total costs. Google says it will continue development on fast action, rapid script changes and complex physics, but has not announced dates for those improvements.
Key Questions
How long can Gemini Omni 1.2 Flash videos be?
Google says the model generates native clips from three to 15 seconds, with durations selectable in one-second increments. Version 1.1 had a stated maximum of 10 seconds.
What is the listed price for 720p video?
Google lists 720p video at $0.07 per second, which it says is 30% below the price for Gemini Omni 1.1. The announcement does not give prices for other resolutions.
Can users edit a generated clip without starting over?
Google says users can make changes with follow-up prompts, including requests to alter a background or lighting. Independent information about how reliably edits preserve the rest of a clip was not included.
Where is Gemini Omni 1.2 Flash available?
Google says it is rolling out through the Gemini API and Gemini for business, and also points to Google AI Studio, Flow and Google Vids. It has not published a platform-by-platform or regional rollout schedule.
Have the claimed quality improvements been independently verified?
Not in the supplied announcement. Google’s claims about reference following, consistency, speech and lip-sync were not accompanied by independent benchmarks or failure-rate data.
Primary source: Google AI · via ThorstenMeyerAI.com
