Google Veo 3.1 takes AI video making up a notch with sound and new control tools

New Delhi: Google has upgraded its AI video generator with the launch of Veo 3.1, the newest version of its filmmaking model that now brings richer audio, better realism, and finer editing control to creators. The company said that the update builds on Veo 3, released in May, and makes it easier for users to create lifelike videos with natural sound and stronger prompt adherence.

The new version has been rolled out across Google’s Flow video editor, Gemini app, and enterprise platforms such as Vertex AI and the Gemini API. Flow, Google’s AI video creation tool powered by Veo, has already seen over 275 million videos generated since its debut earlier this year.

Veo 3.1 adds generated audio to all major video generation features in Flow, including "Ingredients to Video,” "Frames to Video,” and "Extend.” Google said this gives users "more narrative control” and makes scenes sound more natural.

With "Ingredients to Video,” users can upload multiple reference images to define characters, objects, and visual styles. Flow then uses these ingredients to generate a complete scene that closely matches the user’s vision.

Another feature, "Frames to Video,” allows creators to provide a starting and ending image, and Flow will automatically create a seamless transition between them. It’s a useful tool for cinematic shots or artistic video bridges.

The "Extend” tool helps users create longer sequences by generating new footage that connects to the final moments of an existing clip. Google said that with Veo 3.1, these extended scenes now come with rich, AI-generated audio to maintain continuity and atmosphere.

Google is also adding new editing options within Flow for users who want more creative flexibility. Two key tools are being introduced: Insert and Remove.

"Insert” lets users add new elements to any video scene, such as realistic background details or entirely imaginary objects. Flow now automatically handles shadows and lighting to make these additions look like they naturally belong in the shot.

Meanwhile, the "Remove” feature, which will roll out soon, will allow users to delete unwanted objects or characters. Flow will then reconstruct the surrounding visuals to make it look as if the element was never there.

Google said that Veo 3.1 brings enhanced realism with truer-to-life textures and stronger adherence to text prompts, improving the quality of both visuals and audio. These refinements make AI-generated videos look more consistent across scenes, an area that earlier versions struggled with.

According to Google’s announcement, Veo 3.1 also features improved narrative understanding, meaning it can better interpret the story a user wants to tell. This helps maintain logical flow and continuity between clips, especially in longer projects.

Along with Flow, the Veo 3.1 model is also available for developers via the Gemini API, and for enterprises through Vertex AI, expanding its reach beyond consumer creators. Google confirmed that a lighter, faster variant called Veo 3.1 Fast has been launched for quicker video generation tasks.