HN Explorer

Ask HN: Any good (free?) TTS article readers?

6 points by terabytest ↗ HN
I really enjoy pasting articles into ChatGPT and having its voice model read them out to me. It sounds way better than other TTS systems I’ve tried, but it’s cumbersome.

Are there any interesting products in this space?

7 comments

[ 3.5 ms ] story [ 30.1 ms ] thread
viraptor
I'm listening to articles through Google assistant ("read this page to me") and find it perfectly acceptable.
hoothin
I'm using Microsoft azure TTS, which you can use their API within the free quota, and this is a Python program I wrote to generate speech and srt subtitle files from text utilizing their API. Feel free to refer to it. https://github.com/hoothin/AIApplications
solardev
Have you checked your operating system? On macOS, in Chrome and Safari, you can select any text on a page and just go to Speech -> Start speaking and it'll read it using the built-in TTS from your OS.

You might want to change the default voice (under macOS Settings -> Accessibility -> Spoken Content) to "Siri - Voice 4", which will sound much better than the default Samantha.

If you click on "Manage voices", you can also download other voices.

--------

If you want better voices, you can use the OpenAI voices via desktop apps like https://goodsnooze.gumroad.com/l/voices or Chrome extensions like https://chromewebstore.google.com/detail/readme-text-to-spee...

kinow
I use the Read Aloud Firefox extension to read newspaper articles and the wandering inn stories. Works well for me. In the configuration there are free and premium voices. I enable Google Translate voice and increase the speed a bit.

Main repo (website there too, or search on Firefox extensions): https://github.com/ken107/read-aloud

ppedra
I also use this one. If you have a free tier AWS account, you can use the "amazon polly" voices. The neurals voices are very good. If you only use it to read articles to yourself (i.e. a few articles a day and, on some days,none at all), I don't think you'll reach the limit (which is reset monthly). And if you do, switch to normal voice (not neural) and wait for the month to end