AI avatar in Microsoft Teams: Tavus PAL joins, Sume makes clips
Tavus PALs can join Microsoft Teams calls by invite address or Teams URL. Sume renders 4-60 second 16:9 avatar clips you share into a meeting.
For an avatar that takes part in a Microsoft Teams meeting, Tavus's September 12, 2026 changelog says its PALs can join Teams calls. Sume does not join meetings: it renders a finished avatar video, up to 60 seconds, that you can share into a Teams meeting as a file.
Tavus facts are from its changelog; Sume's from Generate avatar video, read 2026-09-30.
How does a Tavus PAL join Teams?
The changelog says PALs can now join Microsoft Teams calls in addition to Google Meet and Zoom. You invite the PAL's @tavusinvite.com address to a calendar event that has a Teams link, or pass a Teams URL when creating a conversation.
What does Sume give me for a meeting?
A clip. Avatar videos turn a ready avatar into a script-driven talking video, accepted when the estimated duration is 4-60 seconds. aspect_ratio supports 16:9 (default 9:16), and resolution is 720p. The result can include a public media.sume.com video artifact you can download and play or attach.
| Question | Tavus PAL (changelog) | Sume avatar video (docs) |
|---|---|---|
| Joins a Teams call | Yes, by invite address or Teams URL | No |
| Responds during the meeting | Conversation with the PAL | No; script is fixed before render |
| Output | A participant in the call | A downloadable video file |
| Length | Not a clip | 4-60 seconds per video |
When is a clip enough?
If you want a presenter to deliver a known message, such as an intro or announcement, play a 16:9 clip when the agenda reaches it. If attendees need to ask questions and get answers, a clip cannot do that; the live route is the fit. The same split is covered in real-time avatar vs video avatar API.
How do I make the 16:9 clip?
Submit to POST /v1/avatar-1.0/talking-video with an avatar_handle, a script, and aspect_ratio set to 16:9, with an Idempotency-Key header. Poll the job until it is terminal, then fetch the artifact. For scripts longer than a minute, split them into several jobs.
Sources
Related posts
More in Use cases
- AI change text in image: an edit call with a reference
To change wording inside a picture, send it as an input_references image with a prompt giving the new text. Sume lists edit-capable models.
- AI music for one section of a video: no duration field
Sume's Music 1.0 rejects duration and duration_seconds. Ask for the length in the prompt, then loop and fade the track to fit one section on a Timeline.
- AI object remover from a photo: mask_url on GPT Image 2.5
To remove an object from a photo through the Sume API, send the photo and a public mask_url to POST /v1/images with a GPT Image 2.5 model, then prompt the fill.
- AI product color changer: colorways through an image edit API
To recolor a product photo, send it as an input_references image with a prompt naming the new color. Sume's edit calls take up to 10 images per request.
Written by Sume