Parse a document
Parse one uploaded document on the server. API-key and cli:inference OAuth callers send only the multipart file field, with no model, trust tier, or parse mode. Returns parsed text, document_type and token_count. Token count is a model-independent UTF-8 byte estimate, not model-specific usage. Billing and spending caps are enforced before processing; one document-parse unit is charged. Content is not persisted; responses are no-store.
Authorizations
Org API key as Authorization: Bearer z_sk_...
Body
Request payload for parseAttachment. Examples use placeholder-safe values and do not include real secrets or private prompt content.
Upload one file for server-side parsing using an API key or CLI OAuth credential. No model, trust tier, or parse mode is accepted.
Multipart binary file upload to parse.
Response
OK
Schema for AttachmentParseResponse values returned by or accepted by the Nito gateway.
The document_type field on AttachmentParseResponse.
"example"
The text field on AttachmentParseResponse.
"Say hello."
Model-independent estimate: UTF-8 byte length of parsed text divided by four, rounded up. Not a model-specific tokenizer count or billing usage.
x >= 01

