Skip to main content
Maybloc is a speech-to-text API for recordings with more than one voice. You upload a file and get back a transcript in which every speaker is separated, overlapping speech is kept, and every word has a start and end time. The model is calico-1-flash. It costs $0.006 per minute, billed by the second, and transcribes recordings of 20 seconds to 3 hours in 45 languages. The base URL is https://app.maybloc.com/api/v1.

​
What does the API return?

Every completed transcript has four parts:
  • Speakers: each voice gets a stable id such as speaker_1 and its total speaking time.
  • Turns: one speaker talking without a break, with the text and the time of every word.
  • Overlaps: the start and end of each moment when two or more people talk at once, and which turns overlap.
  • Activity: when each speaker is audible, including short sounds with no words.
See what a transcript result contains for the full shape.

​
How do I make my first request?

Quickstart

Upload a recording, transcribe it and read the result in four requests.

Authentication

Create an API key and send it as a bearer token.

Add-ons

Redaction, summaries, speaker names, sentiment, translation and more.

Webhooks

Get a signed request when a transcript finishes, instead of polling.

​
How much does it cost?

$0.006 per minute for the model, plus a per-minute price for each add-on you switch on. A failed or canceled transcript is not charged. See pricing for every price and a worked example.

​
Is there an OpenAPI specification?

Yes. The OpenAPI 3.1 document is at https://app.maybloc.com/openapi.json. The API reference is generated from it.
Last modified on October 26, 2026