Zoom realistic avatars need your camera on: a recorded avatar clip
Zoom's realistic avatars use your live camera in meetings. If you want an avatar to deliver a weekly update while you are away, here is the recorded route.
Zoom's Avatars blog describes avatars that stand in for you in live meetings. For realistic and stylized avatars the camera must be on, because the avatar follows your face using on-device face detection, and the page says a liveness check runs when you create one. It lists up to 10 realistic or stylized avatars, or 25 cartoon avatars, and notes that a visible watermark marks them as avatars. I read the page on 2026-10-04.
That is a live-presence feature. If your need is an update that plays later while you are on a plane, a recorded clip is the closer tool.
Live avatar versus recorded avatar
A live avatar needs you present: you move, it moves. A recorded avatar needs a script. With Sume you write the words, pick the avatar, and get an MP4 you can post or send anywhere. Nobody needs to be on camera at the moment of delivery.
| Question | Zoom realistic avatar | Sume talking video |
|---|---|---|
| Needs your camera on | Yes, for realistic and stylized | No, script in |
| Created from | Your face, with a liveness check | A prompt, a profile or a public photo |
| Marked as an avatar | Visible watermark, per Zoom | You disclose it yourself |
| Use | Live meetings | A finished 720p MP4, 4 to 60 s |
| Cost basis | Not stated on the page | Per second, from $0.184 |
How to send a weekly update without being there
Write the update as a script of 120 to 150 words, which comes to about a minute of speech. Call POST /v1/avatar-1.0/talking-video with your avatar handle, aspect_ratio: "16:9" for a desktop audience or "9:16" for a phone, and optional captions. A 45 second clip at standard costs 45 x $0.184 = $8.28, and at plus 45 x $0.245 = $11.03.
Because the estimated duration is checked before the job starts, a script that would run past 60 seconds fails fast, and you can trim it.
A consent note
If the avatar looks like a real colleague, tell viewers it is an avatar and get that person's consent first. Zoom marks avatars with a visible watermark for that reason; with Sume you decide how to disclose it in the caption or the first line of the script.
Sources
Related posts
More in Sume Avatar 1.0
- Introducing Sume Avatar 1.0
Sume Avatar 1.0 is a multi-agent orchestration system as a single avatar model.
- Avatar Face Swap API (Beta): apply an avatar face to a video
Avatar Face Swap 1.0 is a Beta Sume endpoint that applies a ready avatar's face to a short public source video. Required fields, limits, and polling.
- Avatar video previews: approve the first frame before rendering
Create an avatar video preview to get first-frame stills, regenerate them if needed, then call generate-video on the preview id to render the final video.
- Multi-scene avatar video API: build one video from ordered scenes
Send ordered video_inputs instead of one script to compose spoken and silent scenes into one 4-60 second avatar video. Scene fields, rules, and limits.
Written by Sume