Mary sounds quite good. Apparently if you know how to write input files in the correct format you can get it to speak much more like a real human by including stress/speed/tone/etc. variations too.

>http://espeak.sourceforge.net/

Can't confirm how well espeak works, since I didn't try it, but I have tried pyttsx on Windows. (pyttsx is a Python library which uses SAPI5 on Windows and espeak on Linux.)

Worked okay for me. Not tested a lot though. Here's my simple Python snippet for trying pyttsx on Windows:

http://code.activestate.com/recipes/578839-python-text-to-sp...

(A commenter on the above ActiveState recipe said my pyttsx recipe worked fine for him with espeak on Crux Linux.)

and more details on the same here:

http://jugad2.blogspot.in/2014/03/speech-synthesis-in-python...

"The pyttsx library is a cross-platform wrapper that supports the native text-to-speech libraries of Windows and Linux at least, using SAPI5 on Windows and eSpeak on Linux."

Also, check out the synthesized train announcer's voice in my blog post above - somewhat eerie and cool :)

And a similar small trial of the reverse problem - speech recognition, using the Python 'speech' library:

http://jugad2.blogspot.in/2014/03/speech-recognition-with-py...

Perfect moment to mention my little project.

http://simulationcorner.net/index.php?page=sam

I have to admit, that the quality is not good. But it also works on embedded systems.