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

# Deepgram

> Use Deepgram nova-3 to transcribe meeting bot recordings on Zoom, Google Meet, and Teams: full config reference with defaults, diarization, keywords, and example payloads.

Deepgram's `nova-3` model transcribes the bot's recording after the meeting, with diarization, smart formatting, and paragraphing on by default.

## Enable

```json
{
  "meeting_link": "https://meet.google.com/abc-defg-hij",
  "bot_name": "Notetaker",
  "recording_config": {
    "transcript": {
      "provider": { "deepgram": { "model": "nova-3", "language": "en" } }
    }
  }
}
```

CLI: `meetstream bot create <link> -t deepgram`.

## Configuration reference

| Field             | Type    | Default  | Notes                        |
| ----------------- | ------- | -------- | ---------------------------- |
| `model`           | string  | `nova-3` | Only supported value         |
| `language`        | string  | `en`     |                              |
| `punctuate`       | boolean | `true`   |                              |
| `smart_format`    | boolean | `true`   | Formats dates, numbers, etc. |
| `diarize`         | boolean | `true`   | Speaker separation           |
| `paragraphs`      | boolean | `true`   |                              |
| `numerals`        | boolean | `true`   |                              |
| `filler_words`    | boolean | `false`  | Include "um", "uh"           |
| `utterances`      | boolean | `true`   |                              |
| `utt_split`       | number  | `0.8`    | Utterance split sensitivity  |
| `detect_language` | boolean | `false`  |                              |
| `keywords`        | array   | —        | Boost domain terms           |
| `search`          | array   | —        | Search terms                 |
| `tag`             | array   | —        | Request tags                 |

## Fetch the transcript

After `transcription.processed` fires (or `GET /bots/{bot_id}/detail` shows the `transcript_id`):

```bash
curl "https://api.meetstream.ai/api/v1/transcript/<TRANSCRIPT_ID>/get_transcript" \
  -H "Authorization: Token <YOUR_API_KEY>"
```

Add `?raw=true` for Deepgram's raw response. Want live output instead? Use [Deepgram Streaming](/guides/transcription-recordings/providers/deepgram-streaming). Compare engines: [Transcription Providers](/guides/transcription-recordings/providers/transcription-providers).