Gift box insert: a QR code to a setup video you can swap later
Put a QR code on a gift-box insert that opens a setup video. Make the clip on Sume, caption it for sound-off, draw the QR in code, and point it at a redirect.

Print the QR code on the insert from a short link that you own, and let that link redirect to the setup video. Make the video on Sume, burn captions in for people who open it with the sound off, and draw the QR code with a library, not an image model. When the video changes, you change the redirect. The card in the box stays the same.
Gift recipients are the people who did not choose the product. They open a box, find a device or a kit, and have no order page to read. A short setup clip behind a QR code on the insert is the cheapest way to meet them, but a printed code cannot be edited. Every choice below comes from that one constraint.
Step 1: make the setup clip
Start from what you already have. If a person filmed the setup on a phone, video trim cuts the useful span out of the file for a flat $0.02 per job, and Timeline 1.0 joins several spans into one MP4 at $0.10 per output minute, both read from the docs on 2026-10-10. The live rates are in GET /v1/catalog.
If you have no footage, the Sume Format catalog has a sume-product-usage-demo slug and a sume-mobile-app-ugc slug. Read the Format first with GET /v1/formats/sume/{slug}: it returns the description and the io profile, so you know what input the run expects before you spend anything. See the Format catalog for the call shape.
Keep the clip short. A recipient who scans a code on a card wants step one in the first seconds, not a brand film.
Step 2: caption it for the sound-off case
A person scanning a code on a sofa often has the phone muted. Sume's standalone video captions endpoint burns text onto a finished clip. If the clip has speech, the job transcribes it, and you can send script_text to keep the words exact. If the clip is silent, speech-to-text fails with caption_no_speech, and the documented fix is to send your own cues, each with text, start and end.
One accepted caption job is $0.20 for a video of up to 60 seconds, per the same page (read 2026-10-10).
curl -X POST https://api.sume.com/v1/video-captions \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-H "Idempotency-Key: insert-setup-captions-001" \
-d '{
"video_url": "https://media.sume.com/artifacts/artf_demo/setup.mp4",
"style": "slam",
"cues": [
{"text": "1. Charge the base for two hours", "start": 0, "end": 4},
{"text": "2. Press and hold the top button", "start": 4, "end": 8}
]
}'Step 3: print a link you can still change
Do not encode the video URL itself. Encode a short link on your own domain, such as a path per product, and have your web server redirect it to the current video. Then a better clip, a corrected step or a new firmware screen only needs a new redirect target. This is a plain web practice, not a Sume feature, and it is the reason the print run is safe.
Draw the QR code with code. An image model can paint something that looks like a code and does not scan, which is why the ticket art and Muse Image posts both draw it after the art. Error correction level Q survives a small scuff or a fold line better than the default.
import qrcode
from qrcode.constants import ERROR_CORRECT_Q
def make_insert_qr(short_url: str, path: str) -> None:
qr = qrcode.QRCode(error_correction=ERROR_CORRECT_Q, box_size=12, border=4)
qr.add_data(short_url)
qr.make(fit=True)
qr.make_image(fill_color="black", back_color="white").save(path)
make_insert_qr("https://go.yourshop.test/box/setup", "insert-qr.png")What each step costs and which tool does it
Every Sume URL you pass to Timeline must already be a media.sume.com artifact or asset, so import outside files first with POST /v1/media-imports, described in Media inputs.
| Step | Sume tool | Docs rate | Note |
|---|---|---|---|
| Cut phone footage | Video trim | $0.02 per job | One span per job |
| Join spans | Timeline 1.0 | $0.10 per output minute | Hosted files only |
| Burn captions | Video captions | $0.20 per job, up to 60 s | Silent clip needs cues |
| Draw the QR | Your own code | Not a Sume job | Redirect target stays editable |
Before you print
Print a proof on the real card stock and scan it with two different phones. Open the redirect on mobile data, not your office network. Watch the clip once with the sound off. Then print the run.
If the same box ships in several languages, make one clip and one redirect per language and print the matching code on each insert. Keep the redirect names boring and per product, so support staff can find the right one.
Sources
Related posts
More in Use cases
- Golf Instructor Lesson Videos: Where AI Helps and Where It Does Not
A golf coach should film the real swing. Sume handles the finishing: trim at $0.02, captions at $0.20, a music bed at $0.13 and a $0.10 per-minute render.
- Gym class schedule promo: a silent Wan clip and three timed cues
A gym class schedule video on Sume: one 6-second Wan 3.0 480p clip and one caption job with three timed cues. $0.575, no speech-to-text, read 2026-10-10.
- Holiday-market walking tour teaser: one 30-second clip for $1.88
A local guide can post a 30-second holiday market tour teaser from Wan 3.0 for $1.88 at 480p. Where 30 seconds fits on Google Business and Shorts.
- India Rule 3(3): metadata, identifier and no-removal for AI media
India's IT Rules now require tools that make synthetic media to label it and embed permanent metadata. What the rule text says and what to check on a Sume file.
Written by Sume