How can I split and feed large Zoom recordings into OpenAI Whisper model?

Hi @AlexisN 

Good question.

Perhaps there’s a Zap app in the Video Category that can be used: https://zapier.com/apps/categories/video

@Troy Tessalone ,

Thank you so much for your reply!  I looked through the 77 apps in that category and didn’t see anything that seemed to be a solution.  A few seemed like they might, but when I looked at their zap options, they didn’t offer file splitting.

Thanks for your help!

Hey there @AlexisN, welcome to the Community! 🎉

I did some checking and found we’ve got an existing feature request open for the 25MB file limit to be increased. If that limit was increased it would solve the issue for you here so I’ve gone ahead and added your vote for that. I can't make any promises around when or if it might be added, but we’ll definitely notify you by email once it is.

I also looked at the supported apps in that category too and couldn’t see any that mentioned being able to split or compress the audio files. Are the audio files always over 25MB or is it only occasionally over that?

If it’s only occasionally then for now you could manually split the audio files using a third party audio file splitter app. Then you’d save the split files into in an app like Google Drive for example. And have a separate Zap that triggers when each of the split audio files is added into a certain folder in Google Drive which then sends them over to OpenAI to generate a transcript.

Alternatively, are there any settings in Zoom you could tweak to reduce the quality of the audio files? Just thinking that should help to reduce the file size. That said, I also found that Zoom has an audio transcription feature which might be worth exploring? 

If you do manage to get any of those suggestions working, or find an alternative solution in the meantime please do let us know. Really keen to ensure you’re all set! 🙂

Dolby.io can be used to compress audio files and possibly get them under the 25mb limit, but this method isn’t reliable for most purposes as the resulting file size isn’t completely predictable. 

The Zoom audio transcription feature is only available to Business accounts, it’s not even included with Pro, so that makes it unavailable to the majority of users. We REALLY need to lift the 25mb Whisper limit.

I understand where you’re coming from @Salamander44.

I have added you as another interested member to the open feature request to increase the file size limit beyond 25 MB. While we can’t provide an exact timeframe for the feature’s implementation. We will let you know via email once there’s an update.

I appreciate your patience and understanding.

I am also interested.

Hi @nexuslux,

I have added you as to the feature request as well. We will keep you in the loop via email once the feature has been implemented.

Thanks!

i’m also interested in this feature

Hi @bwm! 👋 

I’ve added your vote to that feature request as well. Can’t make any promises around when or if it will be added but we’ll be sure to email you once it is. 

I’d also like to upvote this feature request.

Hi @Nate Imhoff,

I have added you as another interested member to the feature request. We will let you know via email or on this thread once the feature has been implemented.

Thank you for your patience and understanding.

@ken.a  @SamB  but isn’t whisper also limited to token size? becasue i compressed an audio file of a 40 minute podcast and it had 13 mb but couldn’t be processed

Good question, @Leismann! 

I couldn’t see any specific token limits outlined for Whisper though I did spot that for files that are being split up it will only consider the final 224 tokens of the prompt. So maybe that same limit applies for files that have been compressed?

Alternatively, I wonder if it might be an issue with the rate limiting for Whisper that you ran into. It looks like they have different Request Per Minute and Requests Per Day allowances for their different plans. Do you think that might have been the issue? If so, if you replay the failed run of the Zap a bit later on it should in then replay successfully and process the file.

Do let me know if you’d also like to be added to the feature request here and we’ll be happy to submit your vote! 🙂

I’d like more than 25MB -- that;s way too small!!  Can you add me too! 

Thanks for reaching out here @LegalZapperGuy!

Your vote has been added to the feature request now. We’ll be in touch if/when that limit is increased. 🙂

Interested as well!

Thanks for letting us know @Artgeek!

I’ve added your vote to the feature request to have the file limits increased. We’ll be sure to contact you and share an update the thread here once it’s implemented. 🙂

I'm encountering a similar issue. It's crucial to support file sizes larger than 25MB. Audio files that are one hour long are typically larger than 25MB. The ability to process meeting note audio files would be an extremely useful feature for Zapier users utilizing OpenAI's Whisper.

Hi @Stone! 👋

Really appreciate you reaching out here to share your support for this feature request. 

We don’t have an ETA on when files larger than 25MB will be supported but I’ve added your vote for it which will help to increase it’s priority. And allow us to notify you by email once the file limits have been increased.

Hello, I am also very interested in this feature. Thank you

Yes, we want this feature as well… need to automatically transcribe and summarize an hour long meeting - trying to use Zapier to push out the whole process.

Hi. I’m also interested in this feature. Thank you