180 seconds of AI video: cost by model at 480p, 720p and 1080p

Generating 180 seconds of video costs between $6.75 and $255.88 on Sume depending on model and resolution, and resolution alone moves a Seedance 2.5 bill from $48.36 at 480p to $255.88 at 1080p. Aspect ratio does not change the price in this table; resolution does.

Three minutes is the Reels ceiling that Metricool reports Instagram's guide lists, and it is a convenient round number for a budget. The numbers are Sume billable amounts per output second times 180, read from the pricing package on 2026-10-03. Most catalog models cap a clip at 15 seconds (Seedance 2.5 and Wan 3.0 go to 30, Omni takes 3 to 10), so a 180-second piece is always stitched from several clips in the Timeline.

The grid

Seedance prices are token-based, so resolution changes pixels per frame and the bill with it; Wan 3.0 and Omni are flat per second by resolution. The Omni row is 360p, 720p and 1080p; its 4K rate of $0.375 per second would be $67.50 for 180 seconds.

Seedance 2.0 Mini is the lowest per-second 480p price among the Seedance family here, and its clips are capped at 15 seconds, so 180 seconds is 12 clips.

What resolution buys

The jump from 720p to 1080p roughly multiplies a Seedance 2.5 bill by 2.5 and a Wan 3.0 bill by 2. Whether that is worth it depends on where the video plays: a phone feed hides the difference more than a landing page does.

A hybrid saves money without much visible cost. Render the hook and the product close-ups at 1080p (30 seconds on Seedance 2.5: $42.65) and the rest at 480p (150 seconds: $40.30), for $82.95 instead of $255.88.

Add the finishing

Footage is not the finished video. Add narration, captions, a Timeline render and music and a 3-minute piece picks up about $1.15: $0.13 narration at 2,700 characters, three $0.20 caption jobs, three $0.10 Timeline minutes and a $0.125 track. That number is the same at every resolution.

So finishing is a rounding error at the top of the table and a visible fraction at the bottom.

Reading the grid for a client quote

When a client asks for a price on three minutes of video, the honest answer is a range with the resolution named. For vertical social video, 720p is the usual default (Auto create controls default to 720p), so the 720p column is the one to lead with: $22.50 on Wan 3.0 or Omni, $34.02 on Seedance 2.0 Mini, $68.04 on Seedance 2.0 and $104.00 on Seedance 2.5.

Quote retakes separately. The grid assumes every second you generate is used. If half again as much footage is generated and discarded, multiply the footage line by 1.5 and leave the finishing line alone.

What Sume does and does not do

Sume prices every output second at the resolution you request and reserves it on submit. It does not generate three minutes in one call; stitch clips in the Timeline, which accepts up to 1,800 seconds of output. Confirm rates in the catalog.

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 = [
    ('Seedance 2.5 1080p seconds (hook, product)', 30, 1.42155),
    ('Seedance 2.5 480p seconds (the rest)', 150, 0.268677),
    ('Narration characters', 2700, 0.0000475),
    ('Caption jobs', 3, 0.20),
    ('Timeline minutes', 3, 0.10),
    ('Music tracks', 1, 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}")