Home → Blog → How to set up transcription without writing code
How to set up transcription without writing code
You do not need a developer to transcribe recordings automatically. Every automation platform can send an HTTP request with a file attached, and that is the entire requirement. Here is the setup for n8n, Make and Zapier, plus the one mistake everybody makes with the file field.
What the platform needs
The recipe is identical everywhere:
- Endpoint: https://voicesscribe.com/v1/audio/transcriptions, method POST.
- An Authorization header with the value Bearer followed by your key.
- A form body with a binary field named file and a text field model set to whisper-1.
The response is JSON with a text field — that is what the rest of the scenario consumes.
n8n
Use the HTTP Request node: method POST, URL https://voicesscribe.com/v1/audio/transcriptions, authentication via header. Under Body pick Form-Data and add two parameters — model with the value whisper-1, and file of type n8n Binary File pointing at the binary property from the previous node, usually named data.
The classic failure here is sending the file as text. If the response is 400, nine times out of ten the file field carries a string instead of a binary property.
Make and Zapier
In Make, use the HTTP module with the Make a request action. Set the body type to multipart/form-data, map the file field to the binary output of the previous module (a Google Drive file, an email attachment), and set model to whisper-1.
In Zapier, Webhooks by Zapier with a Custom Request works. The file is passed as the attachment from the previous step and the authorization header is added manually. Zapier caps attachment sizes, so compress long recordings before they reach the step.
Where the text goes next
Three connections that take about half an hour to build:
- Inbox to spreadsheet. An email with an attachment arrives, the file is transcribed, the text lands as a new row.
- Cloud drive to knowledge base. A new file in a watched folder becomes a page with its transcript.
- Telephony to CRM. A call recording is transcribed and attached to the deal as a note.
If the scenario processes batches, add a concurrency limit: platforms love firing everything at once and running straight into the rate limit.
Try it on your own recordings. Sign-up takes a minute, and the free minutes are enough to judge the quality.
Get a free API keyFrequently asked questions
Do I need a dedicated integration module?
No. A generic HTTP node, which every automation platform ships, is enough.
What is the maximum file size?
25 MB. Many no-code platforms cap attachments lower still, so compressing to opus first is a good habit.
How do I get subtitles instead of plain text?
Add a response_format field set to srt or vtt and the response comes back as a ready subtitle file.
Why does my scenario return 400?
Check that the file field is passed as binary data rather than a filename or a link. That is the most common cause.
Related reading
- How to build a batch transcription pipeline for an audio archive — Transcribing thousands of recordings without losing any: a work queue, concurrency, retries, resuming after a crash and keeping the bill under control.
- How to turn voice notes into text automatically — Turn phone and recorder voice notes into a searchable knowledge base: a synced folder, a watcher script, a note template and tips for messy field audio.
- Transcription API errors and what they actually mean — A walkthrough of transcription API status codes: why 401, 400, 413, 429 and 502 happen, how to fix each one and which failures are worth retrying.