HeyGen photo avatar renders without a reference look (Aug 2026)
HeyGen's August 2026 change lets an Avatar V photo avatar render from the photo alone; motion_prompt still needs a reference look. Sume's photo path, compared.
HeyGen's changelog for August 2026 says an Avatar V photo avatar can now render directly from the photo when its group has no eligible digital-twin or curated reference look, instead of the request being rejected. The fallback applies when reference_look_id is omitted and no eligible look can be picked automatically. motion_prompt still needs an animation reference. On Sume the photo path is two steps: create an avatar from the photo, then render by handle.
What did HeyGen change?
The entry, titled "Avatar V Photo Fallback", says video creation no longer rejects a photo avatar that lacks an eligible reference look. It renders from the photo. It adds that motion_prompt still requires an animation reference for Avatar V photo avatars, and that you should supply an eligible reference look when you need prompted body motion or hand gestures.
So the change removes a rejection; it does not add motion control for photo-only avatars.
| Question | HeyGen | Sume |
|---|---|---|
| Photo as the source | Photo avatar; render from the photo if no look | input.type: "photo" with a public HTTPS image_url |
| Reference look needed to render | No, as of August 2026 | No look concept; the avatar handle is the reference |
| Prompted body motion | Needs a reference look | Not a field on talking-video |
| Reuse | Avatar group | avatar_handle, stored without @ |
How does the photo path work on Sume?
Post the photo to POST /v1/avatar-1.0/generate with an avatar_handle and input.type: "photo". The image_url must be a fetchable public HTTPS image: localhost, private-network, non-HTTPS URLs and non-image responses are rejected before generation.
Poll the job until it completes, then render with POST /v1/avatar-1.0/talking-video using that handle.
curl -X POST https://api.sume.com/v1/avatar-1.0/generate \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: avatar-image-001" \
-d '{
"avatar_handle": "reference_presenter",
"input": {"type": "photo", "image_url": "https://example.com/reference.png"}
}'Which should I pick?
The choice comes down to the motion you need and how you reuse the avatar.
- You want a quick render from one photo and no motion prompt: both routes do this.
- You need prompted hand gestures on a photo avatar: HeyGen's entry says that needs a reference look.
- You want the avatar reused by a short name in code: Sume stores a handle normalized without
@. - Check that the avatar is
readybefore submitting a render on Sume, or the request returnsavatar_not_ready.
Sources
Related posts
More in Comparisons
- HeyGen pause tag vs a Sume silence scene in an avatar video
HeyGen supports break tags up to 5 s on professional voice clones. Sume has no inline pause tag documented; you add a silence scene inside video_inputs instead.
- HeyGen Studio Templates API vs a reusable Sume avatar request
HeyGen can now create templates from videos with POST /v3/templates. Sume has no avatar-video template object; the request body plus a handle is the template.
- HeyGen Video API 5-15 s at 768p vs Sume avatar video 4-60 s
HeyGen's new Video API makes 5-15 second clips at 480p or 768p from a prompt. Sume's avatar video renders scripts of 4-60 seconds at 720p. Which fits your job?
- HeyGen scene-scoped edits vs fixing one scene in a Sume avatar
HeyGen's Video Agent can edit one scene and keep the rest. Sume has no scene-edit call: structural changes need a new preview; only quality changes late.
Written by Sume