Veo, developed by Google DeepMind and launched in 2024, is the underlying video-generation model powering Google Flow, also covered in this directory, but independently accessible through Google's developer API and Vertex AI as well, for teams that want to build video generation directly into their own products rather than using Google's own consumer app.
The model has built a strong reputation specifically for physical realism, water, cloth, and object motion that follows real-world physics more convincingly than several competing models, alongside longer native generation durations than some rivals support, differentiating it technically rather than only through interface design.
A free tier covers limited use, and paid plans start around $20 a month for more generations, with separate enterprise API pricing through Vertex AI. For a developer or enterprise team that wants to integrate a physically realistic video-generation model directly into their own product via API, rather than using Google's consumer-facing Flow app, Veo's independent API access addresses that directly.
