PodNotes
PodNotes copied to clipboard
Ability to fetch transcripts from Apple Podcasts
Apple Podcasts recently released support for transcripts in Podcasts . It would be great if we can fetch that from the API .
Proposed Workflow :
- Give an option (toggle) to fetch transcripts
- This can be used to get summary of a podcast episode if needed with other Quickadd LLM models
https://www.apple.com/newsroom/2024/03/apple-introduces-transcripts-for-apple-podcasts/
Hi @Tanzeel159, thank you for suggesting this!
I'd love to have this feature, but can't find any API from iTunes that lets me grab the transcriptions.
A while back I started to implement this feature on a branch, but it would cost a lot to transcribe the episodes. Would love to query some repository that just exposes it.
Oh okay so they haven't implement it yet. I will keep track and share if something is useful . Thanks for the effort though for trying to implement it yourself .
For info: on IPhone Apple podcast app, I tried to copy the full transcript but when paste it somewhere is just taking the beginning of the copied text
Added transcription feature in #101 via OpenAI's Whisper
I've been using it all day! Works super well. LMK if you find it useful as well, or if you have any ideas/suggestions. Closing this for now :) Just open a new issue if you have an idea / dm me.
@chhoumann can you explain how to use it?
@chhoumann can you explain how to use it?
Please see https://podnotes.obsidian.guide/transcripts/
@chhoumann is giving me [Error transcribing chunk 0] [Error transcribing chunk 1] [Error transcribing chunk 2] although I put an API key
@zhampu please check the developer console for more information. You can access it with Ctrl+Shift+i, or Cmd+Shift+i if you are on Mac. It should display the error messages there.
Did you add billing for the OpenAI account and project?
BIlling is OK
plugin:podnotes:10235 POST https://api.openai.com/v1/audio/transcriptions 429 (Too Many Requests)
The error 429 means "Too many requests," so it sounds like you've reached a rate limit for how much you can send to the OpenAI API. Waiting for a few moments resolves the issue.
any way to fix this?