Brewery label that comes alive: Kling 3 first and last frame, 84 cents

Give Kling 3 the plain label as the first frame and your finished label as the last frame, and ask for 6 seconds. On Sume that job is $0.84 silent or $1.26 with sound, at either 720p or 1080p. Because the last frame is fixed, the clip lands exactly on your approved artwork, which is the property a printed label needs.

How the request looks

Sume's /v1/videos takes frame_images with frame_type set to first_frame or last_frame. Kling 3 lists both on Sume, with 16:9, 9:16 and 1:1 aspect ratios and 4 to 15 seconds. Use 1:1 for a can or bottle label shot square, or 9:16 for a story.

import os
import requests

body = {
    "model": "kling-3",
    "prompt": "The hand-drawn hops on the label unfurl and glow, then settle",
    "duration": 6,
    "resolution": "1080p",
    "aspect_ratio": "1:1",
    "generate_audio": False,
    "frame_images": [
        {"type": "image_url", "image_url": {"url": "https://example.com/label-plain.png"}, "frame_type": "first_frame"},
        {"type": "image_url", "image_url": {"url": "https://example.com/label-final.png"}, "frame_type": "last_frame"},
    ],
}
r = requests.post(
    "https://api.sume.com/v1/videos",
    headers={"Authorization": f"Bearer {os.environ['SUME_API_KEY']}"},
    json=body,
    timeout=60,
)
print(r.status_code, r.json().get("polling_url"))

Price by length and sound

fal's Kling v3 Pro page, read 2026-10-08, lists $0.112 per second with audio off and $0.168 with audio on. Sume bills the provider list price times 1.25, rounded up to the cent per job.

Making the end frame work

Small choices decide whether the animation looks intended.

• Make both frames the same size and aspect ratio as the request, 1:1 in the example.
• Keep the background identical in both frames so the motion happens in the artwork, not the room.
• Draft a 4-second silent version first at $0.56.
• Add the clink and pour sounds in your editor; at $1.26 versus $0.84, the audio option is a convenience, not a requirement.

Limits

Kling 3 on Sume has no reference-image input, so it will not keep a logo consistent beyond the two frames you give it. Long text on the label may blur in motion; test with your real artwork.