Google Vids pauses its custom avatar rollout until further notice

Google Vids will build a custom AI avatar from an image you upload, instead of only from a menu of Google's stock presenters or a build-your-own screen. Google announced the change in a Workspace Updates post dated November 2, 2026, then paused the rollout of the feature the next day.

The pitch is narrow and practical. If you make training videos, onboarding walkthroughs or internal explainers, you have probably wanted a presenter who looks like someone at the company rather than a stranger from a stock library. Vids previously made you pick one of Google's characters or assemble a face from a build-your-own menu that let you set traits such as age, gender and race, as Android Authority described it. The new option skips the assembly step: hand Vids a photo, and it turns that photo into a presenter you can reuse.

What Google Vids added

Google Vids will let users create a custom avatar from an image they upload, not only from scratch. That is the whole change. The Workspace post frames it as "greater options for customization and more precise creative control by uploading your own image, in addition to the ability to build a custom avatar from scratch."

Three things happen in order, per Google's own description. You choose an image. Vids refines it inside the editor. The finished avatar is saved to the project so it can carry through a video.

The image does not have to be one sitting on the machine you are using. Google says you can pull it from a local computer, from Google Drive or from Google Photos. The refine step is the editor's own, and Google lists three examples of it: cleanup of backgrounds, adjustments to lighting, and festive decorations.

The "save for reuse" detail matters more than it sounds. A custom avatar that is stored at the project level appears consistently across the scenes of one video, which is the difference between a presenter and a parade of near-misses. Google's post puts it simply: "Custom avatars are saved to the Vids project, so they can be reused consistently within a video."

How to build a custom avatar from a photo

Android Authority walked the workflow, and the steps run inside the Vids editor:

1. Create a new Google Vids project and choose Create AI videos.
2. In the right pane, go to the Create tab and click Avatar.
3. Click Design a custom avatar.
4. On the next page, choose Upload image, which now sits alongside the existing templates.
5. The following page shares guidance on the best image to use. Click Upload to insert your image.
6. Vids processes and enhances the image. From there you can use it to create a new clip.

One constraint comes with the shortcut. If you build an avatar from an uploaded photo, you choose from the existing voice options rather than a voice made for that specific face, Android Authority reports. The image gives you the look; the voice is still picked from Google's list.

Android Authority adds a practical upside over Gemini's separate avatar tool: because each avatar starts from one photo, you can build several looks for different clips by uploading different images, instead of writing custom instructions to steer a single character.

Where the image can come from, and what Vids does to it

The three sources Google names cover most of where a company photo already lives. A headshot on a laptop, a team picture in Drive, an old portrait in Photos. You do not need to export anything first or run the file through a separate tool, because the picker reaches into all three places.

The refine options are light retouching, not a rebuild. Cleaning up a background, evening out lighting and adding decorations change how the image reads without changing who it is. Google does not describe a way to change features, age or expression of the person in the photo, so the avatar stays tied to the picture you supplied.

What the feature does not include

Two absences are worth stating plainly, because both affect how a team rolls this out.

There is no admin control. Google's post says it directly, under the heading for administrators: "This feature does not have an admin control." A Workspace administrator cannot switch the uploaded-image avatar on or off for a domain, so once the rollout reaches an account, anyone with Vids access can use it.

Google's announcement does not state a consent or permission requirement. The Workspace post describes the upload paths, the refine step, the project-level save, the rollout status and the editions, and it says nothing about needing the photographed person's agreement before turning their picture into a presenter. That silence is the finding: Google does not say. The advisory firm Adaptd, in its own write-up of the feature, recommends the opposite as a matter of policy, and its advice is to get written permission before turning anyone's photo into an avatar. That is an outlet's recommendation, not a Google rule, and the two should not be confused.

The voice limit is the third boundary. Google's post does not mention it; Android Authority's walkthrough does, reporting that an uploaded photo pairs with the existing voice options.

Which editions get it, and when

Google lists the eligible editions in the same post. On the business side that is Business Starter, Business Standard and Business Plus; on the enterprise side, Enterprise Starter, Enterprise Standard and Enterprise Plus. Google also names Enterprise Essentials and Enterprise Essentials Plus, Nonprofits, and Individual, plus Education Plus and the Education add-ons Google AI Pro for Education and Teaching and Learning, and the AI Expanded Access add-on. For personal Google accounts, the feature reaches users on Google AI Pro and Google AI Ultra.

The timing is unsettled. Google paused the rollout on November 3, 2026, the day after it announced the feature, and moved the rollout pace for both tracks back to TBD: "Rollout for this feature is paused. We will post an update when it resumes." The dates the post carried before the pause are gone from it, so no start date is current for either release track. Android Authority, writing on November 3, 2026, reported the feature was already reaching Workspace users and would soon reach Gmail users with AI Pro or Ultra subscriptions; that sequencing is Android Authority's reporting, not a line in Google's post.

What it means for creators

No one can switch the feature on yet. The rollout was paused on November 3, 2026 and every listed edition sits at TBD until Google resumes it, so the piece of the announcement a creator can act on today is the pause, not the upload button.

When it does resume, the feature lowers the cost of a consistent on-screen presenter. For a training library, the same face can front a whole series of clips without a reshoot, which is the case Vids is built for. Because it rides on the entry Business Starter plan, a small team does not need a top-tier contract to try it.

The catch is governance rather than technology. With no admin switch and no consent language in the announcement, the decision of whose face becomes a company avatar lands with whoever has the Vids tab open. A team that wants a rule will have to write its own, and treat it as policy rather than as a setting Google provides.

Google Vids custom avatars are paused for every listed edition as of November 3, 2026; when the rollout resumes, eligible users open a Vids project, choose Create AI videos, then Avatar and Design a custom avatar to upload an image. Google's help page walks the avatar workflow at support.google.com.