Three hook variants for one avatar ad: $3.31 on Standard

Three 6-second hook variants for the same avatar ad cost $3.312 in total on standard (3 x 6 s x $0.184). Rendering the winning script as a full 30-second video on max is a further $16.50. You decide on the expensive tier only after you have seen the hooks.

Draft cheap, finish expensive

Avatar Video bills per second of video, and the tier changes only the rate. standard is the fastest path, plus is the default, and max is the highest quality with slower turnaround. Avatar rates are Sume's charged per-second tier rates from the repo rate card (standard $0.184, plus $0.245, max $0.55 without a product image); confirm them in GET /v1/catalog.

Run the hooks

Hooks are short, so send each as its own script with the same avatar_handle. The minimum is a 4-second estimated duration, so a one-line hook of a few words can still be billed as 4 seconds.

import os, requests

H = {"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"}
hooks = {
    "a": "Stop scrolling: this one change saved us ten hours a week.",
    "b": "Nobody tells you this about editing video.",
    "c": "We cut our edit time in half. Here is how.",
}
for key, text in hooks.items():
    r = requests.post("https://api.sume.com/v1/avatar-1.0/talking-video",
        headers={**H, "Idempotency-Key": f"hook-test-{key}-v1"},
        json={"avatar_handle": "ad_host", "script": text,
              "quality": "standard", "aspect_ratio": "9:16"}, timeout=60)
    r.raise_for_status()
    print(key, r.json()["data"]["request_id"])

Limits and gotchas

• A hook rendered on standard does not change when you re-render the full ad on max: the voice and avatar are the same, the render quality differs.
• If you want to approve the first frame only, create an avatar video preview first and pass quality at generate-video; preview stills are tier-independent.
• Add a product image only to the final: it raises the per-second rate (for example $0.194 instead of $0.184 on standard).