▣ INGESTION KIT

Convert Speech to Text & Subtitles API

What it does

The Audio to SRT API turns an audio or video file into subtitles or a transcript. Send a public url or a base64-encoded file in audio, then set format to srt, vtt, txt or json. MP3 works as well as video files, so you can send recordings as they are.

For captions, max_chars (20 to 200, default 84) and max_seconds (1 to 30, default 7) control how long each cue is, so subtitles stay readable in video players and editors. The json format returns the full text, the duration in seconds, timed segments with start and end times, an average confidence score, and billed_minutes. Set word_timestamps to true to get every word with its own start time, end time and confidence.

Teams use the Audio to SRT API as a video to SRT and subtitle generator for YouTube and course videos. Others add closed captions for accessibility, make podcasts and recorded calls searchable, and store meeting transcripts. It works from Python, Node.js or any language that can send an HTTP request. AI agents can call it through ApyHub MCP.

For plain transcription of WAV recordings, see the Speech to Text API. To turn text back into audio, use the Text to Speech API.

POST

Transcribe audio or video

https://api.eu.apyhub.com/callable-labs/convert-speech-to-text-subtitles/v1/transcribe

QUICKSTART

GUIDE

Quickstart

Transcribe an audio file from a public URL and return subtitles/transcript in the default JSON format.

curl -X POST "https://api.eu.apyhub.com/callable-labs/convert-speech-to-text-subtitles/v1/transcribe" \
  -H "apy-token: $APY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "audio": {
      "url": "https://assets.apyhub.com/samples/sample-speech.mp3"
    }
  }'

What you'll get back

Returns a JSON object with text (the full transcript), duration (audio length in seconds), segments (the timed transcript segments), confidence (average token confidence from 0 to 1), and billed_minutes (minutes charged, rounded up).

{
  "text": "Hello and welcome to the show.",
  "duration": 61.2,
  "segments": [
    {
      "start": 0.32,
      "end": 2.4,
      "text": "Hello and welcome to the show."
    }
  ],
  "confidence": 0.97,
  "billed_minutes": 2
}

TRY ITLIVE · 25 ATOMS

API key

Loading your default key…

The full key is used to call the gateway and stays in this tab — never sent to orbit or saved.

About this endpoint

What it does

Transcribes an audio or video file and returns a structured transcript. You can provide the source either as a public URL or as base64/data URI content, and choose an output format for the transcription.

Request Body

Parameter	Type	Mandatory	Description
audio	Object	Yes	Source file for transcription. Provide one of `audio.url` or `audio.base64`.
audio.url	String	No	Public http(s) URL of the audio or video file. Max length: 2048 characters.
audio.base64	String	No	The file as base64 or a `data:` URI.
format	ENUM	No	Output format. Allowed values: `json` (transcript, segments, timings), `srt`, `vtt`, `txt`. Default: `json`.
max_chars	Integer	No	Longest segment or subtitle cue, in characters. Default: 84. Minimum: 20. Maximum: 200.
max_seconds	Number	No	Longest segment or subtitle cue, in seconds. Default: 7. Minimum: 1. Maximum: 30.
word_timestamps	Boolean	No	Include every word with its start, end and confidence in `json` output. Default: `false`.

Response

Returns a JSON object with text as a string, duration as a number, segments as an array of objects, confidence as a number, and billed_minutes as an integer. The response is the transcription result, including the full transcript text, total audio duration, per-segment timing data, confidence, and billed minutes.

Parameter	Type	Mandatory	Description
text	String	Yes	The whole transcript, punctuated and capitalised.
duration	Number	Yes	Audio duration in seconds.
segments	Object Array	Yes	Transcript segments. Each segment includes `start`, `end`, and `text`; `words` is present when `word_timestamps` is enabled.
confidence	Number	Yes	Average token confidence, 0–1.
billed_minutes	Integer	Yes	Audio minutes this request counts against the plan, rounded up and at least 1.

▣ COMMON ERRORS

Errors any endpoint can return

400bad_request

Required parameter missing or malformed body.

401unauthorized

API key missing, revoked, or not authorized for this service.

429rate_limited

Your plan's per-second rate exceeded. Retry with exponential backoff.

503upstream_busy

Backend temporarily unavailable. Try again in a few seconds.

Similar APIs

Closest matches in the catalog based on this service.

CALLConvert Text to Speech APIApyHub CALLConvert Speech to Text APIApyHub CALLConvert Speech to Text with (Whisper Large-v3)MasterA AI CALLGenerate Video Subtitle API FreeDailyPro

From the blog

Reads that touch this API.

Text to Speech in Python: 5 Ways (Offline, Free and API)Oct 20, 2026 PDF to Speech: Turning Documents Into Audio With One API CallSep 7, 2026 Speech Recognition: Use Cases and SolutionsSep 18, 2026 Converting Video to Audio at Scale: Best Practices for DevelopersAug 5, 2026