Google expands AI avatars in Vids with more languages and custom controls

Google has significantly upgraded the AI avatar features in its video creation tool Google Vids. A post in Google Workspace Updates explains that the changes include more preset avatars, broader language support, and new ways to control custom avatars in generated video clips.

The number of preset avatars has grown from 23 to 53. The new defaults span three visual styles: photorealistic, 3D cartoon, and graphic novel. All presets now use Gemini Audio, which Google says delivers more expressive and conversational speech.

More languages and custom avatar controls

Google has added 16 new languages to Google Vids, bringing the total to 24. The new additions include Hindi, Bengali, Arabic, Russian, Turkish, Vietnamese, and Ukrainian, among others. They join a base set that includes English, Spanish, French, German, and Japanese.

Custom avatars now come with expanded capabilities. Users can choose from more than 30 voices powered by Gemini Audio. More notably, custom avatars can now appear in generated video clips, a feature previously limited to default presets. The underlying technology combines Gemini 3.1 Flash Text-To-Speech and Veo 3.1.

Users can direct their custom avatar in two ways:

• Text prompts that instruct the avatar to walk, talk, or interact with objects
• Uploaded images that place the avatar in specific locations or add branded visuals

Google states that each new video generation preserves the custom avatar’s appearance and voice. No administrator controls are required to access these features.

The update is available across a wide range of Google Workspace plans, including Business, Enterprise, and Education Plus tiers, as well as personal Google accounts with AI Pro or Ultra subscriptions. Users with the AI Expanded Access add-on receive higher usage limits for AI avatars.