Lovart credits reset monthly, top-ups last 366 days: Sume's wallet?

Lovart's own FAQ says monthly plan credits last for the current billing cycle and do not roll over, annual plans reset credits every month, and top-up credits are valid for 366 days from purchase. Sume works the other way round: its plans carry no credit allotment to lose, and wallet top-ups carry a 365-day expiry, one day shorter than Lovart's top-ups. If you hate watching a monthly balance evaporate, the difference is in where the money sits, not in the number of days.

What exactly does Lovart's FAQ say about expiry?

The FAQ text on the page did not include plan prices or credit amounts in what I could read, so this post makes no per-credit price claim for Lovart.

How does Sume handle the same questions?

Sume's public pricing page lists plans by access and concurrency: Free, Pro at $40 a month, Startup at $120 and Scale at $400. It does not list monthly credit allotments. Instead, generation and API calls are charged at each model's published USD rate against one wallet, and the dashboard can start a Stripe-backed top-up. Billing docs describe the dashboard as showing plan status and the available balance.

In the code on main, a completed top-up creates a credit lot with an expiry 365 days after it is credited. The public API exposes the resulting balance at GET /v1/balance and the usage ledger at GET /v1/usage.

Which design fits an uneven month?

A plan whose credits reset each cycle is a use-it-or-lose-it deal: a slow month wastes allowance, a busy month forces a top-up. Top-up packs avoid that on both vendors, since each keeps its credit for about a year. On Sume the subscription fee buys access and concurrency (4 jobs on Pro, 8 on Startup, 20 on Scale), and the spend sits in a wallet that is yours for 365 days.

That favors bursty work: a campaign month, a seasonal launch, a one-off batch of clips. It favors Lovart's plan structure less only if you generate steadily enough to use each month's allowance in full.

What should I do before committing either way?

Write down the number of generations you made in your last three months, the largest and smallest. If the smallest is under half the largest, a reset-every-cycle allowance will waste money in some month. Then buy a small Sume top-up (the minimum is $10), run real jobs, and read GET /v1/usage to see what each job cost before you size a larger one.