Price per concurrency slot: Sume Pro $10, Startup $15, Scale $20

Sume plans sell parallelism, not clips. Pro is $40 for 4 slots, Startup $120 for 8, Scale $400 for 20. Compute the plan fee per slot and queue headroom.

5 min readSume
All posts

On Sume the monthly plan fee buys parallel generation slots, and each clip is billed at its model's published rate. Divide the fee by the slots and you get $10 per slot on Pro ($40 for 4), $15 on Startup ($120 for 8) and $20 on Scale ($400 for 20). Pro is the cheapest slot; Scale is the largest pool. The per-second vendor prices that dominate this month's launch news, such as Google's Veo 3.1 Fast at $0.10 per second at 720p, sit on top of the plan fee.

The table

Concurrency and queue capacity come from the generation admission docs. The default queue capacity is max(3, concurrency_limit x 5), and accepted job capacity is concurrency plus queue. Prepaid top-ups do not raise the processing limit; an admin override can, and the dashboard Concurrency tab shows the effective number.

Plan fee per concurrency slot (read 2026-10-05)
PlanMonthly feeProcessing concurrencyFee per slotAccepted job capacity
Free$01n/a6
Pro$404$1024
Startup$1208$1548
Scale$40020$20120

How to use the number

The fee per slot is not a clip price. It tells you what extra parallelism costs when you are choosing between plans. If a nightly batch of 100 equal clips finishes in 25 rounds on Pro (4 at a time) and 13 on Startup (8 at a time), you are paying $80 more per month to halve the wall-clock time, and your usage bill is identical because it depends on seconds rendered, not on slots.

Two traps follow. First, buying more wallet balance does not buy speed, because concurrency is plan-only. Second, you can submit more than your concurrency: Sume accepts valid jobs as queued until accepted capacity is full, and then returns 429 queue_full. Check generation_limits in the submit response rather than hard-coding this table.

  • Occasional batches: stay on the smaller plan and let the queue absorb the work.
  • Daily deadline batches: price the plan jump against the finish time you need.
  • Read the per-second price from the catalog before you multiply by clips.

Source of the per-second example

The Veo 3.1 Fast figure is from Google's Gemini API pricing page, read 2026-10-05: $0.10 per second at 720p, $0.12 at 1080p and $0.30 at 4K. Sume's own per-second rows are list price times 1.25 and live in the model catalog; this post only prices the slots.

Worked comparison for one batch

Take 100 clips that each run for one slot-round. On Pro the 4 slots process them in 25 rounds (100 / 4). On Startup the 8 slots need 13 rounds (100 / 8 = 12.5, rounded up). On Scale the 20 slots need 5 rounds (100 / 20). The extra monthly fee relative to Pro is $80 for Startup and $360 for Scale. Whether that is worth paying depends on how much a faster finish is worth to you, not on the clip price.

If your batches are rare, the Free plan's single slot and six-job accepted capacity is the tell: you can test, but not run production volume. If your batches are daily and deadline-driven, price the jump to the next slot count against the hours you save. And if your work arrives in bursts, remember that queued jobs are accepted, so a small plan with deep patience can still finish a large list. It just finishes it later.

Sources

Related posts

More in Pricing

All Pricing posts

Written by Sume