Sume Pro vs Startup vs Scale for a 100-clip batch: rounds and queue

For 100 identical 10-second Omni 1080p clips, the usage bill is $188.00 on every Sume plan, because usage is billed per clip at list times 1.25 and rounded up to the cent ($1.88 times 100). What the plan changes is how long the batch takes and how many submit waves it needs. With equal-length jobs, 100 clips need 25 processing rounds on Pro (4 slots), 13 on Startup (8 slots) and 5 on Scale (20 slots). Accepted-job capacity is 24, 48 and 120, so only Scale accepts all 100 in one submit pass.

The arithmetic

Rounds are the clip count divided by the slots, rounded up: 100 / 4 = 25, 100 / 8 = 12.5 so 13, and 100 / 20 = 5. Submit waves are the clip count divided by accepted capacity, rounded up: 100 / 24 = 4.17 so 5, 100 / 48 = 2.08 so 3, and 100 / 120 so 1. These assume every clip takes about the same time, which real renders do not, so treat rounds as a planning figure, not a promise.

What the fee actually buys

If this batch is the only job of the month, the all-in cost is $228.00 on Pro, $308.00 on Startup and $588.00 on Scale. The extra $80 from Pro to Startup roughly halves the rounds; the extra $280 from Startup to Scale cuts them from 13 to 5. If you run this batch once a month and the deadline is loose, Pro is cheaper by far. If you run it daily and a late render costs you a launch slot, the fee is the price of the schedule.

Google's Gemini API pricing page, read 2026-10-05, lists Veo 3.1 Fast at $0.10 per second at 720p, a reminder that model price cuts lower the $188 line while the plan line stays where it is. The more the usage line shrinks, the more the plan fee matters in the total.

• Submit within queue_capacity_remaining, and treat 429 queue_full as a signal to wait, not to retry in a tight loop.
• Retry with the same Idempotency-Key so a retried submit cannot double-bill.
• Read generation_limits from the submit response instead of hard-coding the table.