Google Unveils Veo 3 Fast: A Accelerated Video Generation Model in Gemini Pro Subscription

Veo 3 Fast is a streamlined version of the widely discussed Veo 3 model from Google. According to Josh Woodward, head of Gemini Lab, this new model operates twice as fast while retaining all the essential features of the larger version: 8-second videos are generated in 720p resolution and come with sound. Subscribers to Gemini Pro can create up to three videos per day within the Gemini App, and in the flow neural network editor, each generation costs them 20 credits. Currently, generation is only available from IP addresses in the United States, with additional countries expected to be added soon.

Google is currently working on integrating features into Veo 3 and Veo 3 Fast that will allow video creation based on images, as well as the addition of voiceovers for these videos. Additionally, there are plans to reduce the occurrences of errors where the model fails to execute a given prompt.

I have already tested Veo 3 Fast and shared my best results on the channel «the runaway neural network«. You can always support me by subscribing to the channel.