LLM API pricing: Model API Price List

Every row is a verbatim sentence read from the named source's own page through our fetch service on the date shown, hash-pinned to the stored read (the sha of the page's visible text; the raw page is mirrored to R2).

3 of the 18 named sources that could be read published a usable figure; 15 published none and are recorded as not published, never filled in; 0 could not be read (blocked, dead or unreachable) and count nowhere.

The headline figures (median, minimum, maximum, quartiles) are over ONE figure per source in USD (a source's median where its page yielded several accepted sentences), so N counts sources, never sentences; 300 candidate sentence(s) were refused by a reviewer with the reason recorded.

The fuller derivation this dataset is chartered to grow into (not yet what the figures above are; they are each source's own published figure as read): one row per model per vendor at the vendor's own published input and output price per million tokens, with cached-input and batch rates where published, converted to a cost per typical workload (a 10,000-document extraction, a 100,000-message support deployment) so the token price becomes a monthly bill the buyer recognises; the price per unit of capability tracked over time, which is the only line in this whole subject that falls rather than rises, and the count of vendors publishing versus quote-only

300 candidate sentences on these pages were refused by a reviewer; each refusal and its reason is recorded in the review ledger.