Anam per-second billing and spend cap vs Sume reserve at submit

Anam starts the meter when a session starts and bills per second after included minutes. Sume commits money earlier and differently: at submit it reserves the public USD estimate, captures it on success, and releases or refunds it when a job fails.

Anam facts are from its pricing page; Sume facts from Generation admission, read 2026-10-01.

How does Anam bill?

Billing is by the second. Minutes count from the moment a conversation starts, the timer runs until you end the session, and unused minutes do not roll over. You can set a spend cap in Account Settings to limit charges beyond the included allowance, or set conversation time limits through the SDK.

How does Sume reserve and settle?

Paid generation uses public Sume USD estimates. When a request is accepted, Sume reserves the estimated amount; successful completion captures it; failed jobs and failed queue admission release or refund the reservation where applicable. With too little balance, submit fails with 402 insufficient_credits before provider work starts.

How do I limit spend in Sume?

The docs describe the balance check and reservation, not a per-session cap. For unattended agents, read spend caps for unattended AI agents. For production integrations, prefer async submit with an idempotency key.