Skip to main content
The Recnote API turns clinical conversation audio into structured journal notes. You send audio (a complete recording or a stream of chunks), optionally together with metadata that shapes the output — template structure, language, profession context — and you get back a structured note as JSON, ready to render in your own UI or EHR. The API is fully synchronous: every request returns its result in the HTTP response. There is no dashboard dependency and no callback to wait for.

​
What you can do:

Transcribe audio

Submit a recording and get the transcript back. Single file or chunked.

Generate clinical notes

Produce a structured SOAP-style note from audio or from a transcript you already have.

Bring your own template

Pass a template structure and the note is generated to match your sections.

Multilingual

Swedish, English, Finnish, Norwegian, Danish, German, Turkish, Arabic.

​
Two integration modes

  1. One-shot. You have the full recording. POST /api/v1/notes with the audio file → transcript + generated note in one response.
  2. Chunked. You stream audio while the consultation is ongoing. Open a session, push chunks as they arrive, then ask for the note when done.
Both modes accept the same metadata and return the same note shape.

​
Base URL

https://api.recnote.ai

​
Response shape

A generated note is returned as a JSON object with the clinical sections (reason for visit, subjective, objective, assessment, plan), suggested diagnosis/procedure codes, a combined rendering, and optional tasks. You map these into your own UI however suits your product. See Generate notes for the full schema.