OpenAI Transcription Node Error - Need Help

Hello N8N community,

I’m encountering an error when trying to use the OpenAI node to transcribe an audio file. Here are the details:

Setup

  • Node: OpenAI (Transcribe a Recording operation)
  • Input file: Video file (mp4, 104 KB)
  • File field name: data

Error Message

The error says “Invalid file format. Supported formats: [‘flac’, ‘mp3’, ‘mp4’, ‘mpeg’, ‘mpga’, ‘ogg’, ‘wav’, ‘webm’]”

What I’ve tried

I’m trying to transcribe an audio file that is in the supported format (mp4), but the node is rejecting it. The file size is under the 25 MB limit mentioned in the node configuration.

Questions

  1. Is there something wrong with my configuration?
  2. Does the OpenAI node require any specific encoding for mp4 files?
  3. Has anyone successfully used this node with mp4 files?

Any help or suggestions would be greatly appreciated!

Thank you!

Here are a couple of things you could try if you haven’t figured this out yet. (If you did resolve it, please post a reply to share how.)

  1. Try with a different sample mp4 file to see if you get different results. Some are available here
  2. Analyze the file you are attempting to transcribe using a tool like AConvert

Also, your post didn’t mention which version of n8n you are using. They’re fixing things all the time, so be sure to check an updated version to see if anything has changed.

I’m pretty sure this m4a file is incorrectly encoded.

Where did this file come from, and is there a way to encode the file differently at the source?