n8n-nodes-assemblyai

This is an n8n community node. It lets you use AssemblyAI in your n8n workflows.

AssemblyAI develops industry-leading Speech AI models for transcription and audio understanding, accessible through their API. It provides features like speaker diarization, sentiment analysis, entity detection, PII redaction, prompting (Universal-3 Pro), Medical Mode, and LLM Gateway capabilities for processing transcripts with AI.

Resources

• n8n community nodes documentation
• AssemblyAI API reference
• AssemblyAI API documentation
  • AssemblyAI Speech-to-Text guide
  • AssemblyAI LLM Gateway guide

Quick links

• Installation
• Credentials
• Operations
• Development
• Support & Feedback

Installation

Follow the installation guide in the n8n community nodes documentation.

Credentials

To use this node, you need an AssemblyAI API key.

Prerequisites

1. Sign up for a free account at AssemblyAI and receive $50 in free credits.
2. Navigate to your API Keys page in the dashboard.

Setup in n8n

1. In n8n, go to Credentials → New
2. Search for AssemblyAI API and select it
3. Enter your API key from the AssemblyAI dashboard
4. Click Save

Learn more about API keys in the AssemblyAI documentation.

Operations

File (API Reference)

• Upload: Upload a media file to AssemblyAI for transcription

Transcript (API Reference)

• Create: Start a new transcription job with support for:
  • Speech models: prefer speech_models (priority list, e.g. universal-3-pro,universal-2). Legacy speech_model field is still available for backwards compatibility.
  • Prompting (Universal-3 Pro): pass a prompt (up to 1500 words) to steer transcription style and accuracy
  • Medical Mode: set domain to medical-v1 for specialised medical-terminology accuracy
  • Remove Audio Tags (Universal-3 Pro): strip inline annotations like [laughter], [music], and speaker cues
  • Language detection with expected_languages, fallback_language, code_switching_confidence_threshold
  • Code-switching transcription via the top-level language_codes field
  • Speaker diarization with speakers_expected and speaker_options (min/max speakers)
  • PII redaction with redact_pii_return_unredacted and override_audio_redaction_method (silence)
  • Keyterm prompting to boost recognition of domain-specific terminology
  • Profanity filtering, sentiment analysis, entity detection, content safety, IAB categories, auto-chapters
  • Speech Understanding at create time: translation, speaker identification, custom formatting
  • Summarization fields (summarization, summary_model, summary_type) are still available but deprecated — use the LLM Gateway resource instead.
• Get: Retrieve a transcription by ID
• Delete: Delete a transcription
• List: List all your transcriptions, filtered by status, date, ID range, or throttled_only
• Get Sentences: Get transcript broken into sentences
• Get Paragraphs: Get transcript broken into paragraphs
• Get Subtitles: Export subtitles in SRT or VTT format
• Get Redacted Audio: Get redacted audio URL
• Word Search: Search for specific words in the transcript

LLM Gateway (API Reference)

• Chat Completion: Send chat completion requests to LLMs (Claude, GPT, Gemini, Qwen, Kimi). See available models. Supports:
  • Transcript injection: pass transcript_id to substitute {{ transcript }} in your prompt with the transcript's text
  • Tool calling: provide a JSON tools array; force a specific function via Tool Choice: Specific Function and a function name; multi-turn round-trips via the Tool message role + tool_call_id
  • JSON repair post-processing: opt-in toggle that sends post_processing_steps: [{ type: "json-repair" }] to fix malformed JSON in tool-call arguments or structured outputs
• Speech Understanding: Process speech understanding tasks on existing transcripts:
  • Translation: Translate transcripts into multiple languages
  • Speaker Identification: Identify speakers by name or role
  • Custom Formatting: Apply custom formatting to dates, phone numbers, and emails

Development

To work on this node locally, clone the repository and install dependencies:

git clone https://github.com/gsharp-aai/n8n-nodes-assemblyai.git
cd n8n-nodes-assemblyai
npm install

Build and run with linting:

npm run dev:fresh

The node will be available in n8n at http://localhost:5678

Support & Feedback

Need help or have feedback? Reach out to [email protected].

Stay up to date with the latest features and improvements:

• AssemblyAI Product Changelog
• AssemblyAI Product Roadmap

n8n

n8n is a fair-code licensed workflow automation platform.

License

MIT