Get Llm Pricing
GET /usage/pricing/llm
What an LLM operation of this size would cost. Charges nothing.
READ capped before reading credits. Several tiers have a per-operation
ceiling, and for typical prompt sizes the ceiling is what is actually paid
- so
creditsstops moving when the prompt grows, anduncapped_creditsis what the token arithmetic alone would have produced. A caller that showscreditswithoutcappedgets asked why shortening the prompt changed nothing.
found is false for a tier that does not exist or is not currently
active; credits is then 0, which is not a price.
Authorizations
Parameters
Query Parameters
Pricing tier, e.g. g8_t1. The same value that appears in the llm_tier column of GET /usage/transactions, which is where to find the tier a given operation actually bills at.
Pricing tier, e.g. g8_t1. The same value that appears in the llm_tier column of GET /usage/transactions, which is where to find the tier a given operation actually bills at.
Prompt tokens.
Prompt tokens.
Completion tokens.
Completion tokens.
Header Parameters
Responses
200
Successful Response
object
An OPEN envelope around one report’s answer.
The app wraps every one of these reports in {success, data, meta}, where
meta carries the cache state, the window and - on the agency reports -
failed_org_ids, the orgs that contributed NOTHING to the totals above.
That field is the difference between “this client sent no email” and “we
could not query this client”, so it is forwarded rather than flattened
away.
Open on purpose: data is dict | list in the app and its contents grow
with the product. A closed model would turn every new metric into a
silently dropped one.
object
object
Whether there are more pages
Items per page
Current page number (1-indexed)
Total number of items
422
Validation Error