Ideogram /v2/account/usage line items vs Sume /v1/usage

Ideogram's GET /v2/account/usage returns billed line items in time buckets and needs an API key owned by an organization admin. On Sume the two public reads are GET /v1/balance for the USD balance and GET /v1/usage for usage ledger entries, both scoped to the workspace of the API key.

Ideogram facts are from its usage page, read 2026-10-01.

What does the Ideogram usage endpoint return?

Dense buckets with line items; each carries the billed dollar amount, and per-item products add billed_units (unit, quantity, unit price). bucket_width is 1d (up to 92 days per request) or 1h (up to 168 hours). sources defaults to both api and app so totals reconcile with invoices. The data may lag live traffic by a few minutes, so it is for reporting, not real-time monitoring. Other members' keys receive a 404.

What do Sume's reads cover?

The Public API docs list "Balance and usage" as GET /v1/balance and GET /v1/usage: "Read USD balance and usage ledger entries." In OpenAPI, /v1/usage lists the newest rows and takes limit (1 to 100) plus optional thread_id, run_id, and job_id filters. Rows expose Sume job and request ids and public billing amounts only. On image jobs, usage.cost is the billed USD amount (image docs).

Can I top up through the API?

No. The credits docs say API clients should treat top-ups as dashboard operations; the public Developer API exposes balance and usage reads, not a top-up creation endpoint.

How do I reconcile a month?

Page through /v1/usage with limit, sum the billed amounts, and compare with the balance change. For a CSV route, see export AI API usage to CSV.