A 30% price cut on AI video. So what?
On November 5 Google made Gemini Omni 1.2 Flash generally available. It is a video model. You describe a clip, add a product photo or a short reference video, and it generates the footage, with speech and lip-sync if you ask. The headline is the price. A second of 720p video now costs 7 cents through the API, 30% less than the previous version, Omni 1.1 Flash. Clips grew from 10 to 15 seconds. Neowin has the summary and Google’s announcement lists the rest.
Not a chatbot update, then. This is the model behind product videos, virtual try-ons and short ads. If your business does not make video, the direct answer to “should you care” is no. Two indirect effects are worth five minutes.
Where it shows up for a small business
- Marketing video gets cheap. A 15-second product clip at 720p costs about a dollar of compute. The cost is in the person who writes the prompt, checks the result and approves it.
- You may already have it. Google Vids, part of Workspace, generates 15-second clips with this model at no extra cost on a Workspace subscription. Google says it is coming to Gemini for business as well.
- Your agency’s costs dropped. If you pay someone for social clips or ads, their raw material just got 30% cheaper. That is a fair question to ask at the next invoice.
What the price cut does not change: 15 seconds is a social post, not a company video. Every output also carries Google’s invisible SynthID watermark, so platforms can tell it was generated.
The risk side: cheaper fakes
The same release improved dialogue, lip-sync and multi-speaker audio. That is good news for ad makers and for fraud. A convincing 15-second video of “the owner” asking the bookkeeper to rush a payment now costs a couple of dollars and a few minutes. Google restricts voice references to approved customers. Other tools do not, and the price trend is the same everywhere. The watermark helps YouTube flag the clip. It does not help your bookkeeper on a Friday afternoon.
The fix is procedural, not technical. Any change to payment details or any urgent transfer gets confirmed by a call to a number you already had. We covered the security side of Google’s AI push in Google’s AI patch tool, Gemini 4 Carbon, and whether your stack is ready.
The bigger pattern
AI costs move in two directions at once. The raw generation gets cheaper every quarter. The subscriptions that package it for you get more expensive, see the Microsoft 365 price increase coming in 2026. When you plan next year’s IT budget, treat those as separate lines. One is going down and one is not.
What to do this month
Check whether your Workspace or Microsoft 365 plan already includes video or AI features you pay for and don’t use. Ask your marketing vendor how they price AI-generated clips. Write down the payment verification rule and tell the people who move money. If you want a second pair of eyes on which AI features are worth turning on and which are marketing, book an IT assessment.
FAQ
Does the Gemini Omni 1.2 Flash price drop change my Google Workspace bill?
No. Workspace pricing is separate from API pricing. The change is that Google Vids in Workspace can generate 15-second clips with this model at no extra cost.
What is Gemini Omni Flash compared with other Gemini models?
Omni is Google’s line of models that generate video, with speech, from text and reference images. Flash is the faster, cheaper tier of it. The Gemini models you meet in chat and in Gmail are text and image models, and this release does not change them.
Should my small business use the Gemini API for video directly?
Only with a developer and a real use case, such as hundreds of product clips a month. For one clip a week, Google Vids or your usual marketing tool is the cheaper path, because the cost is in the people, not the API.