Google can now generate your AI videos more quickly than ever

• Google’s Veo 3 Fast generates 720p AI videos twice as fast as standard
• Gemini Pro users can create three videos per day
• Flow Pro users pay just 20 credits per clip

Google's revving up its AI video-making engine Veo 3 with a sped-up version called Veo 3 Fast. Both Gemini Pro and Flow users can access the faster Veo 3 model, which produces 720p videos at more than twice the speed of the previous iteration.

Gemini Pro users get three daily video generations using Veo 3 Fast as part of their subscription. Google Flow Pro users on Google’s creative platform can now generate each Fast video for 20 credits, which is far less than the cost of 720p videos before. Gemini Ultra subscribers get far more for their 250-per-month subscription, of course.

Google Veo 3 Fast doesn't just shorten wait times for impatient video-makers, though. It's part of scaling up Google's AI video infrastructure, allowing more people to use the tools simultaneously without waiting too long.

Josh Woodward, who oversees Gemini and Labs at Google, has also outlined plans in the works to incorporate better subtitle rendering and smoother playback as a result of the improved video creation foundation represented by Google Veo 3 Fast.

Making videos more quickly and efficiently is also useful, as it facilitates the incorporation of additional features. Google’s already testing picture-to-video with voice prompts, where, instead of typing a prompt, you could just say “turn this selfie into a soap opera starring supercomputers” and watch as Gemini gives you a low-resolution clip of "Crays Of Our Lives."

Businesses could also get involved, generating internal update videos or onboarding materials without ever filming anything.

Fast videos

Google Veo 3 Fast also aligns with Google’s strategy to integrate AI tools into absolutely everything it offers customers. They don't need new products every quarter when infusing AI features like Veo into Gemini and Flow attracts new consumers and increases use among existing customers.

Now, Veo 3 Fast isn't perfect, of course. Topping out at 720p might leave resolution snobs a little cold, and better subtitles are not the same as flawless or nuanced subtitles. Words can be misunderstood, and those parenthetical emotional clues, like people speaking sarcastically or using irony, can be challenging for an AI to interpret accurately at times.

Still, it's a lot cheaper to just try again with the model. Three videos a day means you don’t have to be precious about your first idea. You can test and experiment with prompts and edits to get the video how you want it.

Veo 3 Fast is not promising a Hollywood trailer in two clicks, but it is promising that you’ll spend less time staring at progress bars and more time testing your wildest video creation ideas. It's a bit like using models or miniatures to visualize a scene before filming it for real.

And even if a video made with the model won't win an Oscar any time soon, it could speed along the tests of ideas that help filmmmakers work out what they want to do when they turn to human actors and editors for the big, expensive set piece they only have one shot to get right.

You might also like

• You'll be as annoyed as me when you learn how much energy a few seconds of AI video costs
• Sorry Pixel owners, Google Cloud's new AI video feature will debut on another company's smartphones
• Sora needs to up its game to match the new Runway AI video model