> 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.

# AssemblyAI Streaming

> Stream live meeting transcripts with AssemblyAI universal-streaming-english: raw or sentence modes, end-of-turn detection tuning — full config reference with defaults.

AssemblyAI Streaming pushes live transcript chunks to your webhook during the meeting, with fine-grained end-of-turn detection tuning. Modes: `raw` (default) and `sentence`.

## Enable

```json
{
  "meeting_link": "https://meet.google.com/abc-defg-hij",
  "bot_name": "Notetaker",
  "live_transcription_required": { "webhook_url": "https://your-server.com/live" },
  "recording_config": {
    "transcript": {
      "provider": { "assemblyai_streaming": { "transcription_mode": "sentence" } }
    }
  }
}
```

CLI: `meetstream bot create <link> -t assemblyai_streaming --live-transcript-webhook <url>`.

## Configuration reference

| Field                                    | Type    | Default                       | Notes                      |
| ---------------------------------------- | ------- | ----------------------------- | -------------------------- |
| `speech_model`                           | string  | `universal-streaming-english` | Only supported value       |
| `transcription_mode`                     | string  | `raw`                         | `raw` or `sentence`        |
| `sample_rate`                            | integer | `48000`                       |                            |
| `format_turns`                           | boolean | `false`                       |                            |
| `encoding`                               | string  | `pcm_s16le`                   | Only supported value       |
| `vad_threshold`                          | string  | `"0.4"`                       | Voice-activity sensitivity |
| `end_of_turn_confidence_threshold`       | string  | `"0.4"`                       |                            |
| `inactivity_timeout`                     | integer | `300`                         |                            |
| `min_end_of_turn_silence_when_confident` | string  | `"400"`                       | ms                         |
| `max_turn_silence`                       | string  | `"1280"`                      | ms                         |

## The live payload

Same envelope as all streaming providers — `new_text`, running `transcript` buffer, `words[]`, `end_of_turn`, echoed `custom_attributes` — documented in [Live Transcription](/guides/transcription-recordings/live-transcription).

Streaming providers do not produce a post-call transcript. Need both? [Re-transcribe the recording](/guides/transcription-recordings/providers/transcription-providers#re-transcribe-any-recording) with a post-call provider after the call.