v0.0.56

Hyprnote v0.0.56 is out!

I know I say something like this for every update, but we really did ship out some in this version.

A lot of you have experienced a weird 'you' keep popping up on your transcript even when you are not saying anything. This was because we did not have a robust silence detection built. Now we have it! Starting from v0.0.56, you won't have to worry about seeing weird 'you's appearing during your recording.

You can now go to your settings' general tab and select the languages you usually speak during meetings. The Whisper model will now only transcribe the languages you selected. No more foreign languages randomly appearing on your transcript. If you leave spoken languages blank, it will just pick whatever language it finds to be closest to.

You can now select which microphone you want Hyprnote to listen to!

We had some users not being able to properly give Hyprnote permission to listen to either their microphone or system audio. We have implemented more robust audio permission requests during onboarding and in the settings modal. If you had problems with Hyprnote not being able to hear your system audio, go visit the settings tab and try giving it permission to access system recording & audio!

Now in the search/command palette, you can look up a person's name and see the meetings that person was part of!

We have combined several different tabs for viewing different information existing inside Hyprnote (events, notes, people, organizations) into one feature called 'the finder view.' You can open up the finder view easily by clicking the Mac Finder icon on the left sidebar (it replaced the previous calendar icon).

It is currently in beta, and we're still doing some maintenance to make the finder view more reliable and stable, so please let us know if you run into any errors or glitches while using it.

We now offer more robust options for you to connect Hyprnote to your own endpoints. Right now, you can just easily put in your API keys for OpenAI, Gemini, and OpenRouter, and select the model easily. No more need to manually input API URL, base, and models every time you switch to a custom endpoint.

Let us know what more options you would like to see.

We added memory to the chat feature! Now all your chat conversations will be stored and can be viewed again. You can also start new conversations and switch between old conversations you started in that specific session/meeting.

Last but not least, we have rolled out our Pro Plan! It's $35 monthly and $179 yearly. It's still early, but we plan to make the Pro Plan more robust. For more details, please visit our website or docs. But don't worry—thanks to our local-first implementation, we don't gate or limit the number of transcriptions/summarizations you can do, even if you're on the free plan. It's free and unlimited, just like it's been until now.

That's it for v0.0.56! Thank you for your time and support :)