Google unveils Veo 3 Fast, an upgraded AI model for faster video generation

Indian Express

2 days ago

Google has unveiled an upgraded version of its most advanced AI video generation model, Veo 3, that is said to be faster and specifically tailored for enterprises.
The Veo 3 Fast is designed for speed and rapid iteration, Google said in a blog post published on Wednesday, July 30. The AI video generator can be accessed via Vertex AI, Google's developer platform for building machine learning models, from August this year.
The tech giant is also bringing image-to-video capabilities to Veo 3 and Veo 3 Fast. This means that users can create 8-second AI-generated video clips of static images using these tools. They have to upload a source image and a text prompt describing the video they want to create.
Since its launch at Google's I/O annual developer conference in May 2025, Veo 3 has been used to generate a total of seven million videos. The company's enterprise customers have used the AI video generator to create over six million videos since it became available in public preview on Vertex AI in June this year.
'Google Cloud customers around the world are using Veo 3 and Veo 3 Fast on Vertex AI to create professional-quality video content with unparalleled efficiency and creative freedom,' the Alphabet-owned company said.
The upgraded model, Veo 3 Fast, can help marketers respond to trends by quickly generating and testing variations of ad concepts. It can further be used to create AI-generated videos of a company's entire product catalogue based on still images. Veo 3 Fast is also capable of providing animated explainers and training modules in less time, the company said.
To address the risk of deepfakes, Google said that all AI-generated videos created using Veo 3 and Veo 3 Fast will have digital watermarks embedded into every frame using SynthID.
'Veo 3 and Veo 3 Fast are also covered by our indemnity for generative AI services,' the company said. This means that Google will offer protection or cover legal costs if its enterprise customers are sued for copyright infringement, violating intellectual property rights, etc.
Veo 3 became available in India earlier this month. It lets users create high-definition (1080p), eight-second clips with audio in a single step. 'This means you can create scenes with characters that speak with accurate lip-syncing, and sound effects that fit the mood,' Google said.
The video generator has been designed for professional marketing campaigns, product demonstrations, and internal communications. 'Veo 3's native dialogue generation helps businesses connect with an international audience by producing a video once and localising the dialogue for dozens of languages,' the company said.
However, support for Hindi or any other Indian language does not appear to be available yet. It is only accessible for subscribers of the Google AI Pro package, which costs Rs 1,950 per month and is free for students.