Well, you double posted this, but I'll answer here.
Your simplest bet is to not bother piping anything, just use links (a text based browser, http://artax.karlin.mff.cuni.cz/~mikulas/links/
in case you didn't know). If you run it like so:
links -dump file.html
This will strip all HTML and print the plain text file to terminal, so you can pipe that out to a file and off you go.
Hope that helps.