Google just made its Vids video editor a lot more capable. The company announced new features that let users control AI avatars using plain text prompts, telling them how to act in a scene.
With natural language directions, you can instruct an avatar to interact with products, props, or equipment. Google says the system keeps character consistency even as the output gets more dynamic and complex.
Customization goes deeper than just directing actions. Users can tweak a character's appearance, swap out their clothing, and generate entirely new backgrounds, all through text prompts tied to the video's theme.
The update also brings Veo 3.1 support to Vids, along with the ability to export finished videos directly to YouTube. A new Chrome extension lets users record content from their browser as well.
This builds on momentum from last month, when Google added its Lyria 3 and Lyria 3 Pro music models to Vids for generating sound effects and background music.
Why this matters: Google is steadily turning Vids into a full AI production studio. Prompt-driven avatar direction removes another barrier between having an idea and producing a polished video. For marketers, educators, and content creators, this could dramatically cut the time and cost of video production without requiring any acting talent or film crew.
The YouTube export feature is a smart addition that tightens the loop between creation and distribution, keeping users inside the Google ecosystem from script to publish.
