Make an AI avatar from a prompt, traits or a photo: $0.95 a call

Sume Avatar 1.0 creates a reusable avatar from a text prompt, structured traits or a reference image. All three cost the same flat $0.95. See the bodies.

3 min readSume
All posts

You can create an avatar three ways on Sume Avatar 1.0, and each is one POST /v1/avatar-1.0/generate that costs a flat $0.95: a text prompt, structured traits (the props type), or a reference image (the photo type). The result is a reusable avatar handle you pass to later talking videos (Create new avatar).

Pick the input

Avatar creation input union
You haveInput typeFields in the docs
A description in wordspromptprompt
Profile data in your apppropsethnicity, sex, age
A reference picturephotoimage_url (public HTTPS)

Traits and photo bodies

Both bodies come from the docs. The handle may start with @; Sume stores it without the @.

# traits
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-profile-001" \
  -d '{"avatar_handle":"product_host","input":{"type":"props","ethnicity":"Asian","sex":"female","age":28}}'

# photo
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"}}'

What the photo URL must be

Before Sume submits a photo generation it rejects localhost, private-network URLs, non-HTTPS URLs and responses that are not images. Host the picture somewhere public, or upload it first and use the resulting URL.

Budget for a roster

The price is per avatar, not per second, so a roster is a plain multiplication.

Flat Sume price per avatar, read 2026-10-08
AvatarsCost
1$0.95
5$4.75
12$11.40
25$23.75

After creation

Each request creates a job. Poll /v1/jobs/{id}/status until it completes, then fetch /v1/jobs/{id}/result. List avatars with GET /v1/avatar-1.0/avatars.

Send an Idempotency-Key so a retried network call returns the same job instead of a second charge.

Then put the avatar to work

The $0.95 buys the avatar, not a video. Talking videos are billed per second by quality: standard, plus (the default when omitted) or max, at $0.184, $0.245 and $0.55 per second. A 30-second plus video is therefore $7.35 on top of the avatar, and one avatar can be reused for as many videos as you like (Generate avatar video).

Sources

Related posts

More in Sume Avatar 1.0

All Sume Avatar 1.0 posts

Written by Sume