Public APIs
Convert Speech to Text API favicon

Convert Speech to Text API

Documents & Productivity

This API converts any speech to text multiple languages and dialects.

Convert Speech to Text API's website screenshot

About Convert Speech to Text API

Speech to Text API converts WAV audio into plain text. It accepts either a WAV file upload or a public WAV URL, along with an Azure locale (such as en-US), and returns the transcription in a data field of the JSON response. Two endpoints are provided: /file for audio already available in the calling application, and /url for audio hosted elsewhere.

The service performs transcription only, returning a single text string without speaker labels, timestamps, or confidence scores. Requests are limited to 100MB total per call, with a URL-based endpoint recommended for larger files.

It is suited for ingestion pipelines, call-center tooling, note-taking applications, and other workflows that need spoken audio converted into searchable or indexable text, such as archiving voice messages or making meeting recordings searchable.

Key features

  • Transcribes WAV audio files to plain text
  • Accepts either a direct file upload or a public WAV URL
  • Supports Azure locale codes such as en-US for transcription language
  • Returns a simple JSON response with a single text field
  • Two dedicated POST endpoints for file-based and URL-based transcription

Frequently asked questions

What audio format does this API accept?

It accepts WAV audio files, either uploaded directly or referenced via a public URL.

Does it provide speaker labels or timestamps?

No, it returns only a single plain text transcription without speaker labels, timestamps, or confidence scores.

How do I specify the transcription language?

You pass an Azure locale such as en-US in the language parameter of the request.

What is the maximum file size for uploads?

The file endpoint accepts a maximum of 100MB total per request; larger files should use the URL-based endpoint instead.

What does the API response look like?

It returns a JSON object with a single data field containing the transcribed text as a string.

Advertise here

Featured products

  • SerpApi - Search API favicon
  • Screenshot Scout favicon
  • TalorData favicon
  • CoreClaw favicon

Show your product to thousands of developers

· 100k monthly pageviews
· 7k newsletter subscribers

Advertise your product