Swap two people in one clip: H3 Max Recast takes 1 to 4 photos
To change two people in one video, send one reference photo for each. Sume's H3 Max Recast takes 1 to 4 person photos and a 5 to 30 second source video.
To swap two people in one clip, send one reference photo for each person. Sume's h3-max-recast replaces the people in a source video_url with the people from 1 to 4 reference_image_urls, one photo per person, for sources of 5 to 30 seconds at 768p or 1080p (Video Router).
What the request needs
video_url: the source clip, 5 to 30 seconds.reference_image_urls: 1 to 4 public image URLs, one photo for each person.- An optional prompt. The duration is the source length.
- 768p or 1080p.
Check the clip first
Count the people on screen and match your photo count to them. Keep each photo clear of other faces, since one photo is meant to stand for one person. Read the model's capabilities from GET /v1/video-router/models for the exact limits before you submit.
Rights
Only use photos of people who have agreed to appear. Sume's docs describe the model's inputs, not permission to use anyone's likeness.
Sources
Related posts
More in Sume Avatar 1.0
- Tavus Griffin-Lite is invite-only: what can you build today?
Griffin-Lite is a closed Tavus preview. Until you get in, Sume ships rendered avatar clips, lip sync and motion control for talking-head video.
- 12 six-second UGC hooks: Standard, Plus or Max rate math
Twelve 6-second avatar hooks come to 72 seconds: about $13.25 on Standard, $17.64 on Plus and $39.60 on Max at Sume's listed per-second rates, before rounding.
- Wedding save-the-date: a 20-second avatar clip, cost by tier
A 20-second save-the-date spoken by an AI avatar costs about $3.68 on Standard, $4.90 on Plus or $11.00 on Max at Sume's listed per-second rates.
- YouTube Shorts series: a weekly AI presenter episode pipeline
YouTube began rolling out Shorts series on 2026-09-23. A repeatable weekly pipeline for presenter episodes: one avatar, draft at standard, final at max.
Written by Sume