AssemblyAI
Connects to the AssemblyAI transcription API. An agent can transcribe audio from a public URL with optional speaker diarization, sentiment analysis, and entity detection.
Add to an agent node's tools array with _node_type: "tool-assemblyai". See Connector Reference — How connectors are used.
Authentication
| Field | Required | Description |
|---|---|---|
api_key | Yes (or assemblyai_api_key) | AssemblyAI API key, sent as the Authorization header. |
assemblyai_api_key | Yes (if api_key not set) | Used if api_key is not set. |
{
"name": "assemblyai",
"_node_type": "tool-assemblyai",
"api_key": "{{ secret.ASSEMBLYAI_API_KEY }}"
}
Tools
assemblyai_transcribe
Transcribe audio from a URL using AssemblyAI with optional speaker diarization, sentiment analysis, and entity detection.
Parameters
| Parameter | Type | Required | Description |
|---|---|---|---|
audio_url | string | Yes | Public URL to audio/video file |
model | string | No | best (default) | nano |
language | string | No | Language code, e.g. en (auto-detect when empty) |
diarize | boolean | No | Enable speaker diarization |
sentiment_analysis | boolean | No | Detect positive/negative/neutral sentiment per sentence |
entity_detection | boolean | No | Detect named entities (people, places, organisations) |
The tool submits the transcription job, then polls (every 3 seconds, up to 10 minutes) until it completes.
Returns: JSON with text (full transcript), words (array of {word, start, end, confidence, speaker?} with times in seconds), utterances (array of {speaker, text, start, end}), sentiments (array of {text, sentiment, confidence} if requested), entities (array of {text, type} if requested), confidence, duration, language, and provider ("assemblyai").