Memory

Answer a question from hosted memory

Retrieves what the caller's memory holds under scope and has an LLM answer question from it, with citations. The model is chosen by the service and the answer is never streamed, so answer_model and stream are refused. Billed per call plus the answer model's prompt and completion tokens.

post/memory/answer

Request

  • Base URL: https://api.tinyhumans.ai
  • URL: https://api.tinyhumans.ai/memory/answer
  • Auth: HTTP bearer

Request body

scopestring required
questionstring required
question_typestring
question_datestring

The date the question is asked as of

temporalunknown
filtersobject
answer_max_tokensinteger

Capped by the service

answer_instructionsstring
cite_sourcesboolean
include_contextboolean

Also return the retrieved context block

use_pack_idstring

Answer from a previously built context pack

Response

The answer with pack_id, answer, citations, provenance, diagnostics, as_of and, when requested, context_block

Changes

No recorded changes to this operation across all 1 revision of this API.