> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.resemble.ai/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.resemble.ai/_mcp/server.

# Create Pronunciation

`POST https://app.resemble.ai/api/v2/pronunciations`

Upload a single pronunciation with a reference audio file.

### Request Body (multipart/form-data)

| Field   | Type   | Required | Description                                                                                                 |
| ------- | ------ | -------- | ----------------------------------------------------------------------------------------------------------- |
| `word`  | string | Yes      | The word or phrase (2-100 characters). Letters, accented characters, apostrophes, hyphens, and spaces only. |
| `audio` | file   | Yes      | Reference audio file (wav, flac, mp3, m4a, ogg, webm, aac). Duration: 200ms-10s. Max size: 10MB.            |

```bash
curl -X POST https://app.resemble.ai/api/v2/pronunciations \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "word=abemaciclib" \
  -F "audio=@abemaciclib.wav"
```

### Response (201 Created)

```json
{
  "success": true,
  "item": {
    "uuid": "a1b2c3d4-e5f6-7890-abcd-ef1234567890",
    "word": "abemaciclib",
    "status": "pending",
    "active": true,
    "audio_url": "https://...",
    "created_at": "2026-03-12T00:00:00.000Z",
    "updated_at": "2026-03-12T00:00:00.000Z"
  }
}
```

The pronunciation starts with `status: "pending"` while the audio is being processed. Once processing completes, the status transitions to `"ready"` (usually within a few seconds) or `"failed"` if something went wrong. Only pronunciations with status `"ready"` are applied during synthesis.