Google has put a clear price on its new AI video model Veo 2: 50 cents per second of video. That figure turns a fast-moving AI product into something buyers can calculate, compare, and question.
The model, which Google unveiled in December, is built for video generation. The new pricing detail shows how the economics of AI video may work in practice, especially for customers trying to understand what a usable clip might actually cost.
The price of Veo 2 is simple to calculate
According to the company’s pricing page, using Veo 2 will cost 50 cents per second of video. That adds up to $30 per minute or $1,800 per hour.
Those numbers matter because AI video is often discussed in terms of capability. Pricing shifts the focus to usage. A customer does not only ask whether a model can create video; the customer also has to ask how many seconds they need, how many attempts they might make, and how much generated material will be worth keeping.
The per-second structure makes the basic math easy. A short clip has a clear cost. A longer sequence becomes more expensive quickly. The pricing also creates a direct link between experimentation and spending, because every generated second has a price attached.
Why the Avengers comparison is useful but limited
Google DeepMind researcher Jon Barron contrasted Veo 2’s pricing with the blockbuster Marvel movie “Avengers: Endgame,” which had a reported production budget of $356 million, or around $32,000 per second.
That comparison puts the 50 cents per second figure in a dramatic frame. Against a major film budget, Veo 2 looks inexpensive on a per-second basis. But the comparison should not be stretched too far.
Customers are not necessarily going to use every second of Veo-generated video that they pay for. A generated result may need to be replaced, shortened, or discarded. In that sense, the listed price is the cost of output, not automatically the cost of finished material.
There is also a difference between producing a short generated clip and producing a full feature-length film. The source noted that Veo 2 is not likely to generate three-hour “Avengers” epics anytime soon. Google’s announcement highlighted Veo 2’s ability to create clips that are two minutes or more.
The real question is usable output
The most important pricing issue may be the gap between generated seconds and usable seconds. If every generated second were ready to publish, the cost would be straightforward. But the source makes clear that customers may pay for video that does not end up in the final result.
That makes planning important. A customer using Veo 2 may need to think in terms of iteration, not just final runtime. The more attempts needed to get an acceptable clip, the higher the effective cost of the finished video becomes.
This does not make the price good or bad by itself. It means the visible rate, 50 cents per second, is only one part of the budget. The practical cost depends on how much generated material survives the editing process.
- Listed price: 50 cents per second of video.
- Minute cost: $30 per minute.
- Hour cost: $1,800 per hour.
- Key uncertainty: customers may not use every second they generate.
How Veo 2 compares with Sora pricing
The source also points to another price in the AI video market. OpenAI recently made its Sora video generation model available to subscribers paying $200 a month for a ChatGPT Pro subscription.
That comparison is not one-to-one, because the source describes different pricing structures. Veo 2 is presented with a per-second usage price. Sora is described as available through a monthly subscription.
Still, the contrast is useful. It shows that AI video tools are being packaged in different ways. One approach makes each second of output directly visible in the bill. Another approach ties access to a subscription tier.
For users, that difference changes how costs are understood. A per-second price makes each clip feel measurable. A monthly subscription frames cost around access. The better fit depends on how often a customer generates video, how much they keep, and how predictable their usage is.
What the pricing reveals about AI video
Veo 2’s price gives the market a concrete reference point. The model is no longer only a technical announcement from December; it now has a published cost that users can evaluate.
The clearest takeaway is that AI video pricing can look affordable in small units while still becoming significant at longer durations. 50 cents per second sounds compact. $30 per minute is easier to feel. $1,800 per hour makes the scale impossible to miss.
The pricing also highlights the difference between generating video and finishing video. A model can produce clips, but the customer still has to decide what is useful. The final value depends not just on how much the model can create, but on how much of that output is good enough to use.
For now, Veo 2’s published rate makes AI video easier to budget and easier to compare. It also gives creators, companies, and customers a sharper question to ask before generating: how many seconds are worth paying for?