While video models often “hallucinate” unwanted details – such as extra fingers or unexpected objects – Veo 2 generates them less frequently, making the results more realistic.
Veo 2 was guided by our commitment to safety and responsible development. We have intentionally measured the growth of Veo availability so that we can support identify, understand and improve the quality and security of the model, while slowly rolling it out through VideoFX, YouTube and Vertex AI.
Like the rest of our image and video generation models, Veo 2 outputs include an undetectable SynthID watermark that helps identify them as AI-generated, helping to reduce the risk of misinformation and misattribution.
Today, we’re bringing modern Veo 2 capabilities to our Google Labs video generation tool, VideoFX, and expanding the number of users who can access it. Visit Google Labs to sign up for the waiting list. We also plan to expand Veo 2 to YouTube Shorts and other products next year.