What happened
Google just made AI video generation both cheaper and longer in a single release. The company shipped Gemini Omni 1.2 Flash, a model that turns a written prompt into a video clip, and in the same update it pushed the maximum clip length to 15 seconds while cutting the price by 30%, down to $0.07 for every second of 720p footage.
If you've never paid for AI video before, here's the plain-English version. You describe a scene in words — a barista pouring latte art in a sunlit café, camera slowly pushing in — and the model hands you back a finished video file. Most tools used to charge per generation, which turned budgeting into a guessing game. Per-second pricing works more like a taxi meter: the number only climbs while the wheels are actually turning. Google confirmed the change on its developer blog, and The Decoder crunched the same numbers on day one.
Why 15 seconds beats four clips stitched together
Fifteen seconds is about the length of a TV ad's opening hook, or one complete beat of a TikTok or Reels clip. Shorter generations used to force you into editing work: make four 4-second pieces, cut them together by hand, and pray the lighting matches. Fifteen seconds means a product reveal, a recipe step, or a talking-head intro can come out in one pass.
It changes the writing, too. With four seconds you describe one action. With fifteen you can describe a beginning, a middle, and an end — which is a lot closer to how people actually tell stories.
What seven cents a second really costs
Run the math and the number stops sounding abstract. A full 15-second clip at 720p costs less than a takeaway coffee. Ten drafts cost about what you'd drop on lunch. And a hundred clips — roughly a month of social content for a small brand — still come in well under the price of a single hour with a freelance editor.
That's the real headline: AI video is drifting from special budget line toward coffee money. If you used to generate three options and pick the winner, you'll now generate ten and pick the best one.
What it means for you
Here's where it actually touches your week, whether or not you'd ever call yourself a 'creator'.
At home
Birthday invitations, a moving-announcement clip for the family group chat, a 15-second 'here's how to reset the router' video for your parents. Small stuff, sure — but it used to mean either paying a stranger or losing an evening to an editor. At pocket-change prices, you'll just make it.
At work
If your job involves anything with a dashboard, a product update, or a training doc, you now have a video option that doesn't require a media team. A 15-second explainer of a new policy beats a wall of email text almost every time.
For your business
This is the sharpest use case. Product demos, seasonal promos, ad variations for A/B testing. Fifteen seconds fits the format the platforms actually reward, and $0.07 a second means testing five different hooks costs less than a sandwich. If you'd rather see what free tools can do before you spend a cent, MyKreaTool keeps a running list of free AI tools worth trying.
For studying
Turn a dry concept into a short animated clip: how a bill becomes law, how interest compounds, how a cell divides. Explaining something visually is a much better test of understanding than rereading your notes, and it's now cheap enough to do for a single exam.
For creative work
Storyboard artists, indie musicians, and writers just got the cheapest sketchpad they've ever owned. You can test a look, a mood, a colour palette in motion instead of describing it in a doc and hoping everyone imagines the same thing.
For extra income
Short-form video is still one of the most in-demand freelance skills on the marketplaces. A 15-second clip that costs pennies to produce can be sold as part of a package — or used to build a faceless channel. The margin is the story here, not the technology.
How to try it right now
You don't need to be an engineer to take this for a spin, though you will need a Google account.
1. Start free. Open Google AI Studio and check the model list for Gemini Omni 1.2 Flash. Playgrounds are where Google parks new models first, so look there before you set up billing.
2. Write a three-beat prompt. Beginning, middle, end in one sentence. Example: 'A runner ties her shoes at dawn, sprints through empty streets, finishes on a bridge as the sun comes up.'
3. Generate once at 720p. A full 15-second clip at $0.07 a second costs less than a coffee — cheap enough that a bad take isn't a crisis.
4. Scale through the API. Once you know what you want, grab a key from Google's developer tools and batch your generations instead of clicking one at a time.
5. Keep a prompt file. Save whatever worked. The cheapest workflow is the one you can repeat next week.
6. Read the fine print. The Decoder's write-up and MLQ's coverage spell out the pricing and length details if you want to check the numbers yourself.
Upsides and what changes
The obvious win is pricing pressure. When Google cuts video prices by 30%, every competitor feels it, and that usually means cheaper tools for you over the next few quarters.
The second change is workflow. Fifteen seconds is long enough that AI video stops being a novelty clip and starts being a building block — a hook, a bumper, an ad unit. Agencies will reorganise around drafts instead of shoots.
Third, the skill that matters shifts. Getting a pretty image out of a prompt was table stakes; now the value sits in pacing and story. Someone who can write a tight 15-second arc suddenly has an edge over someone with a bigger render budget.
Limitations
Honest version: 15 seconds is still short. It won't replace a two-minute product film, and anything with real dialogue, precise lip-sync, or a specific brand logo will need human cleanup. 720p looks fine on a phone but soft on a big screen, and per-second pricing means a habit of hitting 'regenerate' can quietly add up — a long run of failed takes on one clip is real money for a side project. AI video also still trips over hands, on-screen text, and continuity between separate generations, so plan on editing. And treat the model as a tool, not a lawyer: check licensing and disclosure rules wherever you publish, because platform policies around synthetic media keep tightening.
Conclusion
Google's update isn't a revolution — it's a price cut and a longer clip. But both of those quietly change what's worth doing. The bar for 'I could make a video about this' just dropped to roughly the price of a coffee.
One action for today: open Google AI Studio, write a three-beat prompt about something in your own work, and generate a single 15-second clip. You'll know within the cost of that one clip whether this belongs in your toolkit.



Comments 0