I am excited to share WhisperBar with fellow developers or anyone who reads & writes a ton of text when interacting with their AI agent.
For the writing part, WhisperBar does what all other Whisper* apps do but a bit faster and more consistent, with a nice, stable, native macOS and iOS app. But where it really shines is in solving a problem that I'm sure you have encountered too: because of AI, we now have to read and comprehend so much more than before.
Text to voice systems are not available everywhere, but even when they are, they sound so mechanical, and listening to the entire output is even more time consuming.
What Gemini did by turning your research/response into a podcast was genius, and ChatGPT's new voice chat system is also much better (it allows you to have a conversation with the response). However, these two are not available everywhere that I interact with AI (specially with Claude Code and Codex)
With WhisperBar, I can select a piece of text (anywhere, in any app) and press my keyboard shortcut to quickly listen to a short, ~20-second explanation. It does not matter how long the text I have copied is. This allows me to copy the entire response and then hear a summary, or select a section to zoom in on an explanation of that section of the document.
Again, the point is that it is not text to voice, rather, it processes the text and comes up with a natural sounding short explanation.
Oh, and another thing we don't store any data and only use providers with zero-data-retention policies.
It is still a work in progress, and we have about 20 customers right now. It is available on the App Store for both iOS and macOS, with a 7-day free trial at only $4/month (plus Family Sharing). FYI, the WhisperBar iOS app is a bit different (it can record in the background for meetings, etc.).
Let me know what you think :)
Comments URL: [https://news.ycombinator.com/item?id=49457587](https://news.ycombinator.com/item?id=49457587)
Points: 1