n8n integration
Transcribe audio binaries with the Luntrex n8n endpoint.
Point the n8n HTTP Request node at the Luntrex audio endpoint, load models after credential setup, and stream transcripts directly into your workflow.
Configure the audio transcribe node
This guide shows how to authenticate, fetch the model dropdown, and send binary audio from n8n to
/v1/n8n/audio/transcribe.
Integration steps
Create a Luntrex API key
Generate a key in your Luntrex dashboard. You will use it in the n8n credential header.
Set the n8n credential
In an HTTP Request node, set the base URL to your Luntrex API and add
Authorization: Bearer <API_KEY>.
Load model dropdown
After credentials are saved, call
GET /v1/n8n/audio/models to populate the model list
(e.g., openai/whisper-1).
Send binary audio
Use Send Binary Data with Binary Property
data. The controller maps data
to the expected audio upload field.
Execute
Run the node. Responses include
text, status, and the raw provider payload.
Tip: Keep audio files under 25 MB. Use base64 only if multipart upload is not possible.
Parameter reference
- Credential Bearer key for your Luntrex account.
- Model Select from the dropdown after
/modelsloads (formatprovider/model). - Binary Property
data(maps toaudioon the server). - Base64 fallback Send JSON with
audio_base64if you cannot send multipart. - Optional fields
filename,mime_typeto override defaults.
Endpoints
GET /v1/n8n/audio/modelsreturns active audio models for the dropdown.POST /v1/n8n/audio/transcribeaccepts multipartaudiooraudio_base64plusmodel.- Both endpoints require
Authorization: Bearer <API_KEY>.
Example HTTP Request config (n8n)
- URL
https://luntrex.com/v1/n8n/audio/transcribe - Method POST
- Authentication Header
Authorization: Bearer <API_KEY> - Send Binary Data On; Binary Property
data - Body Form field
model=openai/whisper-1
Troubleshooting
- 401/403 Check API key and account credits.
- No models Ensure providers/models marked
model_type=audioare active. - 422 Provide either binary
audiooraudio_base64, and amodel. - Size issues Keep audio under 25 MB or adjust
max_size_mbin model settings.