Google is giving its Vids video editor a big AI glow-up: you can now drop in an avatar with a consistent face and voice, then literally “direct” it to act and speak across multiple scenes using just text prompts. In practice, that means you can create a talking presenter once and reuse them everywhere in your video without weird jumps in appearance or tone.

The feature builds on Google’s Veo 3.1 video generation model, which is already behind Google’s latest AI video tools and is tuned for more realistic motion, better lip‑sync and steadier framing. Google says viewers prefer these Vids avatars five times more often than those on other platforms, which tells you they’re aiming at “good enough for real work” rather than novelty filters.

Inside Google Vids, you pick an AI avatar, type the script, and the avatar will deliver those lines in its built‑in voice, no mic or camera needed. The new twist is control: you can instruct the avatar to walk, pick up objects, or interact with a product or prop just by describing the scene in natural language. Think “Have her stand in a warehouse, put on safety goggles, and point at the emergency exit sign while explaining safety rules” and the system does the rest.

To keep things visually consistent, you can upload reference images—like your office, a store layout, or a product shot—and place the same avatar into each scene while maintaining the same face, style, and overall look. That’s especially handy for training videos, recurring company announcements, or a series of sales pitches where you want one recognizable “host” instead of starting from scratch each time. You can also match avatar voices with Google Vids’ AI voiceover options, so if you mix avatar shots and B‑roll with narration, the voice stays aligned throughout.

Under the hood, all of this runs on eligible Google Workspace and Google AI plans, with Workspace accounts getting promotional access to higher Veo 3.1 avatar limits in Vids before per‑user caps kick in later. The rollout covers a wide range of tiers, including Business, Enterprise, Education, Nonprofits, and Google AI Pro offerings, though for now, the new directable avatars are only available for English‑language Workspace accounts.

For creators, this nudges Google Vids closer to being a lightweight virtual studio: you can script, cast, direct, and publish from a browser tab, and even export videos straight to YouTube once you’re done. It won’t replace full‑blown production houses yet, but for teams that just need clear training modules, internal updates, or simple promo videos with a familiar “face,” it’s a fast, low‑friction way to get there.