> ## Documentation Index
> Fetch the complete documentation index at: https://docs.yolocode.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Voice Transcription

> Convert speech to text with NVIDIA Parakeet

Yolocode provides audio transcription powered by NVIDIA's Parakeet model.

## Endpoints

* `POST /api/transcribe-stream` — Streaming transcription
* `POST /api/transcribe` — Non-streaming transcription

## Supported formats

* WAV
* FLAC
* PCM (raw audio)

## Usage

### JSON (base64 audio)

```bash theme={null}
curl -X POST https://api.yolocode.ai/api/transcribe-stream \
  -H "Content-Type: application/json" \
  -d '{
    "audio": "<base64_encoded_audio>",
    "languageCode": "en-US"
  }'
```

### Multipart form data

```bash theme={null}
curl -X POST https://api.yolocode.ai/api/transcribe-stream \
  -F "audio=@recording.wav" \
  -F "languageCode=en-US"
```

## Parameters

| Parameter | Type | Default | Description |
| - | - | - | - |
| `audio` | string/file | required | Base64 audio (JSON) or file (multipart) |
| `languageCode` | string | `en-US` | BCP-47 language code |

## Notes

* Max request duration: 300 seconds
* Audio format is auto-detected from the data
* The streaming endpoint returns results as they're processed


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.