on october 11th, the tech-media testingcatalog released the blog yesterday, october 10, and was exposedGoogleNew GenerationVincent video model Veo 3.1 The first real-generated samplesenables to generate video clips with a duration of 8 seconds, resolution of 720p and a track。
The media pointed out that a new generation of models, Veo 3.1, had emerged in the cloud platform Vertex AI and Google Vids, and that the first real video samples produced by Veo 3.1 had been released。
Veo 3.1, integrated in Google Vids, already has a core textable video function. Users can enter text alerts and generate video clips with a resolution of 720p over 8 seconds. One of the highlights of this update is the video traverse, which is more contagious than previous generations。
Veo 3.1 made a huge leap in performance from the exposure of comparative test videos such as Cyberpunk hacker robots and volcanic eruptions。
Veo 3.1 understands more thoroughly the nuances of the hints than previous models Veo 3 and produces better images of visual security and detail, such as being able to accurately present the flow sense of lava or more veritable dinosaur form. In addition, 3.1 editions have repaired such issues as the occasional imbalance in the proportion of objects in the previous generation models, and the overall output mass is more stable and accurate。