$regex(eu já sabia,\bjá\b) - returns 0
$regex(eu ja sabia,\bja\b) - returns 1
$regex(eu jáh sabia,\bjá\b) - return 1
$regex(eu jah sabia,\bja\b) - return 0

its a bug or so???

Hello, no it's not a bug, \b by default only checks for ascii, you must use the /u modifier, recently supported by mIRC. It tells the engine to use Unicode to identify characters.

i see... ty verymuch mate

Thanks for your post. I checked your regular expressions at regex101.com and it returned the same results as mIRC. As Wims mentioned, you will need to use the /u modifier in the latest mIRC beta to handle Unicode characters or you can use (*UCP)(*UTF8) at the start of the pattern (see PCRE for more information).

thanks again!