How to read mp3 files from a Google Drive folder

[vanzelleb]

Hi, is there a way to read mp3 files via the Google Drive File reader? It seems to only support text based files. Or can I somehow upload multiple mp3 files to use with the speech-to-text node? I don’t want to manually upload each mp3.

[Gumloop_Bot]

Hey @vanzelleb! If you’re reporting an issue with a flow or an error in a run, please include the run link and make sure it’s shareable so we can take a look.

1. Find your run link on the history page. Format: https://www.gumloop.com/pipeline?run_id={your_run_id}&workbook_id={workbook_id}

2. Make it shareable by clicking “Share” → ‘Anyone with the link can view’ in the top-left corner of the flow screen.
   

3. Provide details about the issue—more context helps us troubleshoot faster.

You can find your run history here: https://www.gumloop.com/history

[Wasay-Gumloop]

Hey @vanzelleb – Yes, you can pass the File output from the Google Drive File Reader node onto the Speech to Text node to read the MP3 file. Here’s how:

You can do the same for multiple MP3 files using a Google Drive Folder Reader node which outputs a List of files from a folder. When you pass the list of files to the Speech to Text node, it would automatically run in Loop Mode for all of them.

Let me know if that works for you.

[vanzelleb]

Hello, I have tried that but the output of the Google Drive Folder Reader is empty despite the folder having .mp3 files in it. You can see the flow here: https://www.gumloop.com/pipeline?workbook_id=u4JJGFfwYpXZFAh7bLXUqf&run_id=fX7VUTwSuxvQyeyJsDsMCH

[Wasay-Gumloop]

Hey @vanzelleb - Could you please try deleting the node, adding it back and trying again? If that doesn’t work could you re-authenticate with drive here please: Gumloop | Settings

Will investigate why it didn’t output files on the first try but it should work. Here’s a successful run link on the same folder: https://www.gumloop.com/pipeline?workbook_id=fkkVvJ33sFE1ohyDrt92xN&run_id=Jvh6Lgp7gtmcRcwaCp9ZNe

[mutty]

I’ve had this issue and I had to format the google drive url to a google drive downloadable url. I eventually ended up using S3 instead. Here’s the code if it helps:

def main(google_drive_url, params):
    # Extract file ID from Google Drive URL
    if 'file/d/' in google_drive_url:
        file_id = google_drive_url.split('file/d/')[1].split('/')[0]
    else:
        raise ValueError("Invalid Google Drive URL format. URL must contain 'file/d/'")
        
    # Construct direct download URL
    google_drive_downloadable_url = f"https://drive.google.com/uc?export=download&id={file_id}"
    
    # Return the downloadable URL
    return google_drive_downloadable_url

[vanzelleb]

After refreshing my Google Drive credentials and creating the node again, I can successfully read the file names. However, the speech-to-text node expects downloadable URLs as input. And as @mutty already mentioned above, this requires the file ID instead of the file name. Maybe in a future iteration, the Google Drive Folder node could output both the file name and the file IDs. For now I am stuck. But it was just a test, so no biggie.

[Wasay-Gumloop]

Hey @vanzelleb @mutty – The Google Drive Folder Reader node outputs the file object itself so you don’t need to use the Use Link option or create downloadable URLs. You can connect the file directly

Example Run Link: https://www.gumloop.com/pipeline?workbook_id=fkkVvJ33sFE1ohyDrt92xN&run_id=Jvh6Lgp7gtmcRcwaCp9ZNe

Note that I’ve a Error Shield around the speech to text node to skip any errors if the file is in a different format that is not yet supported like .mov

[vanzelleb]

It seems you guys updated the speech-to-audio node over the last days. The new version mentions “Audio File” vs. “File URL” as the input. When I use the new version of the node it works like charm. Thanks for helping out.

[Wasay-Gumloop]

You’ve the Use Link option toggled on the node which is why you’re seeing the File URL as input vs Audio File which is the default.