Product launch week content budget: hero video, 5 Reels, 3 carousels

A seven-day product-launch content plan with research, one 30-second hero video, five 15-second Reels and three carousels costs $31.68 on Sume at list rates. The hero video is 56 percent of it, which is the fastest place to save.

The plan is our illustration, not a benchmark, and the quantities are easy to change. Every figure is a Sume billable amount (provider list times 1.25) read from the pricing package and docs on 2026-10-03. Narration assumes 15 characters per second.

The plan and its total

The research line is ten searches of the trending-videos endpoint, which bills $0.10 per accepted call and may serve repeated browse calls from cache. The hero video uses Seedance 2.5 at 720p, $0.5778 per second. The Reels use Wan 3.0 at 720p, $0.125 per second. Carousel slides use Ideogram 4.5 at medium quality, $0.075 per image.

Finishing per video is narration at $0.0475 per 1,000 characters, a $0.20 caption job, a $0.10 Timeline minute and a $0.125 music track.

Unit costs

Each deliverable is built from the same blocks used elsewhere on this site: a clip, narration, captions, a Timeline minute and music for video, and images only for carousels.

Where the money goes

Cutting the hero video from Seedance 2.5 to Wan 3.0 at 720p saves $13.58, which takes the week from $31.68 to $18.10. Doubling the number of Reels adds $11.55. Dropping the music track from every video saves $0.75.

So the levers, from largest to smallest, are the hero model, the Reel count and the finishing stack.

Scaling the plan up or down

A smaller launch drops the research and the carousels and keeps the hero video and two Reels: $22.40. A bigger one doubles the Reels to ten and adds a second hero angle on Seedance 2.5: $61.02. Because every line is a count times a unit cost, you can build either from the unit table.

Two cost details are worth knowing before you set the plan. Trending-videos searches in browse mode may be answered from cache, and the docs say a cached repeat does not re-hit the upstream provider, so an hour of repeated browsing is not ten times the work. And the Timeline rounds up: a 15-second Reel and a 59-second Reel both bill one minute, so if a Reel has room to grow, the render step is not a reason to cut it short.

What Sume does and does not do

Sume produces the assets and bills each job at submit; it does not schedule posts or run your launch calendar. A caveat on the research line: the docs describe each accepted call as $0.10 and cached browse calls as not re-hitting the provider, but do not promise a cached call is free, so budget the full $0.10. For rates, read GET /v1/catalog before quoting.

• Research: $0.10 per trending-videos call.
• Video: per second by model and resolution.
• Images: per image by model and quality.
• Finishing: captions, Timeline, narration, music.

Rerun the arithmetic

The same line items as a short script, so you can change a quantity and see the total. It is plain arithmetic on the rates in this post; it does not call the Sume API.

# Sume billable USD rates, read 2026-10-03. Edit the quantities.
lines = [
    ('Trending-videos searches', 10, 0.10),
    ('Hero clip seconds (Seedance 2.5 720p)', 30, 0.5778),
    ('Reel clip seconds (Wan 720p), 5 x 15', 75, 0.125),
    ('Carousel slides (Ideogram 4.5 medium)', 18, 0.075),
    ('Narration characters', 1575, 0.0000475),
    ('Caption jobs', 6, 0.20),
    ('Timeline minutes', 6, 0.10),
    ('Music tracks', 6, 0.125),
]
total = 0.0
for label, qty, unit in lines:
    cost = qty * unit
    total += cost
    print(f"{label:<46}{qty:>8g} x {unit:<10g} = ${cost:,.2f}")
print(f"{'Total':<46}{'':>8}   {'':<10}   ${total:,.2f}")