Google has made AI video cheaper, longer and far easier to edit. On November 5, 2026, the company announced that Gemini Omni 1.2 Flash, its fast video generation model, is now generally available through the Gemini API, Google Cloud and its business products. The update extends native video clips from 10 seconds to 15 seconds, adds multi-turn conversational editing, and cuts the price of 720p video by 30 percent, to $0.07 per second. For marketers, small businesses, YouTubers and app developers, this is one of the most practical AI launches of the year, because it attacks the three things that have held AI video back: cost, length and control.
The timing is no accident. OpenAI spent the same week rolling out ChatGPT Images 2.6, Meta overhauled its Muse app with real-time voice and avatars, and Thinking Machines Lab shipped Inkling 2 Small, a 594-billion-parameter model. Google’s answer is not a bigger model. It is a cheaper, faster one ordinary people can afford to use every day.
What Gemini Omni 1.2 Flash actually is
Gemini Omni 1.2 Flash is Google’s speed-focused model for generating and editing video with sound. It is multimodal from the ground up: you can give it a text prompt, a still image, an audio clip or an existing video, and it produces a finished video with synchronized audio. The “Omni” in the name reflects that all-in-one input. The “Flash” label marks it as the fast, low-cost member of Google’s Gemini family, built for high-volume work rather than one-off showpieces.
The model reached general availability on November 5, announced by Google DeepMind product manager Anish Nangia. That matters because it is the point at which a model stops being an experiment: developers can build on it through the Gemini API and Vertex AI, businesses can use it in Google Cloud, and consumers will see it across Google’s own apps.
Longer clips and editing you can talk through
The most visible change is clip length. Gemini Omni 1.2 Flash now generates native video clips of up to 15 seconds in a single pass, up from 10 seconds. Five extra seconds may sound modest, but it is the difference between a fragment and a complete thought: a 15-second clip can hold a full product demonstration or a scene with a beginning, middle and end. Short-form video on TikTok, Instagram Reels and YouTube Shorts typically runs 15 to 60 seconds, so one AI-generated clip can now anchor a real post rather than serving as a teaser.
Creators can also draft cheaply and finish in high resolution: generate a rough 360p or 720p draft, refine it, then upscale the finished result to 1080p or 4K.
The second big change is multi-turn editing. With earlier video tools, changing one detail usually meant generating a whole new clip and hoping the rest stayed the same. Gemini Omni 1.2 Flash lets you refine a video through conversation, asking for changes in plain language — make the lighting warmer, swap the background, slow the camera move — while keeping everything else intact. Google says the model holds references more consistently across revisions, so a character’s face, a product’s packaging or a brand’s colour scheme survives the edit instead of drifting with every new version. For advertisers, that consistency matters as much as the extra five seconds, because a video that changes your product’s look every time you edit it is unusable no matter how cheap it is.

The 30 percent price cut changes the maths
Generating 720p video now costs $0.07 per second, down about 30 percent from the previous version. A 15-second clip at 720p therefore costs a little over one dollar, and a minute of finished video costs roughly four dollars at that resolution, before upscaling. For a small business producing a dozen short social videos a week, the weekly video bill drops from a noticeable expense to something close to a rounding error.
Price has been the quiet barrier in AI video: the models have been impressive for over a year, but experimenting was expensive and scaling up prohibitive without a large budget. A 30 percent cut moves AI video from the “special project” budget into the everyday marketing budget, and puts pressure on rivals to justify their own pricing. Note the headline rate applies to 720p output; higher resolutions cost more.
Where you can use it
Gemini Omni 1.2 Flash is available now through the Gemini API, Google AI Studio, Vertex AI and Gemini Ultra, and Google says it is rolling out across its consumer and business products. That includes Flow, Google’s AI filmmaking tool, Google Vids, its video creation app for work, Google Ads, and the Gemini app itself. Google is not launching a standalone video website and hoping people find it; it is putting the same model inside the tools people already use to make ads, presentations and social posts.
For developers, the Gemini API and Vertex AI mean the model can be wired into apps and internal tools: a shopping app could generate a product video from a photo, or a real-estate platform could turn listing photos into a walkthrough clip.
The rollout also reaches Google Ads, which may prove to be the biggest audience of all. Small advertisers who could never afford a video shoot can now generate a 15-second ad from a product photo and a sentence of copy, then iterate by chatting with the model — a direct challenge to the stock-footage and freelance video market.
How it compares with the week’s other launches
The first week of November 2026 was one of the busiest AI launch weeks of the year. OpenAI’s ChatGPT Images 2.6, launched November 4, brought GIF generation and faster image creation to every ChatGPT plan — a consumer play, while Google’s video model targets creators and businesses who need finished clips. Meta’s Muse overhaul added real-time voice and streaming avatars at 25 frames per second with 870 milliseconds of latency: Meta bets on presence, an AI you can talk to; Google bets on production, an AI that makes the video. Thinking Machines Lab’s Inkling 2 Small, a 594-billion-parameter model for coding and knowledge work, is a research story, and Anthropic spent the same week putting Claude into people’s finances with a read-only bank-linking beta. Each lab is carving out different territory; Google’s is affordable video. OpenAI’s other big launch of the week was GPT-6 Cyber, a security model built for authorized vulnerability research.
What it means for creators and small businesses
For independent creators, the most immediate benefit is speed. A YouTuber who needs B-roll or a visual metaphor for a script can now generate a 15-second clip in the time it used to take to search a stock library, and revise it by typing a sentence instead of re-shooting. For a channel publishing daily, that is hours saved every week.
For small businesses, the benefit is access. Video has long been the most effective format on social media and the most expensive to produce. A local shop, a restaurant or a trades business can now turn a photo of a product into a short video ad, test three versions, and keep the one that performs best — all for a few dollars. A 30 percent price cut and conversational editing lower both the cost barrier and the skill barrier at once.
For marketers and agencies, video becomes a testable medium: when a clip costs about a dollar and takes minutes to revise, you can make ten versions of an ad and let the data decide. There are also jobs at stake: stock footage libraries and freelance videographers will feel the pressure, and the premium will shift to taste, storytelling and the ideas behind the video.

The catches, and what happens next
Gemini Omni 1.2 Flash is not magic. Fifteen seconds is still short: a full advertisement or tutorial needs many clips stitched together, and keeping a character or product identical across ten generations is still hard. Fast models also trade some fidelity for speed, and AI video still produces the occasional warped hand or garbled on-screen text, so check every clip before publishing. Several jurisdictions already require AI-generated content in ads to be labelled, and the safest approach is to label AI video clearly. Finally, pricing can change, so businesses building a workflow around $0.07 per second should watch the price list rather than assuming today’s rate is permanent.
Google’s rollout will be the thing to watch over the coming weeks. The model is generally available now, but its appearance inside Flow, Google Vids, Google Ads and the Gemini app will happen in stages, and the quality of those integrations matters more than the raw model. If conversational editing works smoothly inside Google Vids, millions of workplace users will try AI video without ever opening a developer tool. Rivals will respond, as they always do, but for now Gemini Omni 1.2 Flash is the most practical AI video launch of the week: longer clips, editing you can talk through, and a price that finally fits a small budget.
