# Transcribe Audio
URL: /reference/blocks/ai-transcribe-audio
Type: reference
Description: Convert audio files to text using speech-to-text AI
Keywords: transcribe audio, ai/transcribe-audio, ai, block, workflow builder
`ai/transcribe-audio` — Convert audio files to text using speech-to-text AI.

Where it appears [#where-it-appears]

The **Transcribe Audio** block lives in the **AI** group of the Workflow Builder.

Fields [#fields]

| Field       | Type                                   | Required | Default | Description                                                                    |
| ----------- | -------------------------------------- | -------- | ------- | ------------------------------------------------------------------------------ |
| `audioFile` | text (multi-line, supports references) | Yes      | —       | Audio file to transcribe. Supports MP3, WAV, M4A, FLAC, OGG, WebM (max 25 MB). |
| `language`  | text                                   | No       | —       | ISO-639-1 language code (e.g., en, es, fr). Leave empty for auto-detection.    |

Inputs and outputs [#inputs-and-outputs]

Inputs are the configurable fields above. The block produces these outputs:

**Outputs**

Reference an output downstream with `<Transcribe Audio.field>`.

| Name       | Type    | Description                                |
| ---------- | ------- | ------------------------------------------ |
| `success`  | boolean | Whether transcription succeeded            |
| `text`     | string  | Transcribed text                           |
| `language` | string  | Detected language code                     |
| `duration` | number  | Audio duration in seconds                  |
| `segments` | json    | Array of segments with text and timestamps |
| `error`    | string  | Error message if failed                    |

Example [#example]

A minimal configuration for this block:

```json
{
  "type": "ai/transcribe-audio",
  "config": {
    "audioFile": "<PreviousStep.field>"
  }
}
```

Limits and failure modes [#limits-and-failure-modes]

* Required field (`audioFile`) must be set, or the block fails validation before it runs.
* On failure the `error` output is populated; downstream blocks that reference other outputs may not receive values.

Related [#related]

* [All workflow blocks](/reference/blocks)
* [Workflow Builder overview](/workflow-builder-overview)