TL:DR - I personally find MS Zira voice with 75% speed, 100% pitch, to be good enough for most scenarios. Default settings sound horrible, but can be tweaked, that is what I mean.

Okay overall with all things considered, this is currently among the most easy to use TTS (Text to Speech) addons

Pros:
- Easy to use, works fine with a google translated page, easily identifies what you want to read

- You can use Google Wavenet, Amazon Polly & IBM Watson voices with it, if you have API access (all 3 of them have a free tier)

- Has a dark background pop-up that shows the text in a simple readable format alongside speech

Cons:
- Biggest con is that it doesn't use the new Microsoft voices that come with Windows Narrator. It only uses the old Zira, Adam, voices which are of lower quality.

- No way to add a custom Offline voice model. The option of 'Custom Voice is for the online APIs'

- I heard somewhere, that even in offline mode, it does uses some google apis, but haven't checked myself. Given it is recommended by Mozilla officially, I guess its alright for non-security scenarios.

This is a great plug in. Request developer to add open AI TTS support.

Love this feature. Especially since I'm a poor reader.

Hilariously bad. Literally sounds like a backyard full of dogs barking. Not a single word was discernible.

This add-on claims to have "offline" voices which are "never sent to the internet".

However if you block access to the servers it contacts using a firewall:

1) You can see tons of access sent to "support.readaloud.app" and "chart.googleapis.com"

2) The text to speech stops working entirely. It won't read at all and says "Phone not connected". You can't even use the example offline test voice in configurations panel.

How is this even remotely offline if it can't work when not connected to the internet and their servers in particular?

Bro, 2024 with AI and all we got this google roboto voices? fr?

The version in Edge is fantastic, so I figured this one would be as well. No. Just Microsoft Sam quality voices.