Open swanilla opened 4 days ago
same issue here
same, recaptcha seems impossible to bypass at a high rate this days, I've tested a lot of proxies
@ottodriver @swanilla are you guys using recaptcha solvers?
@ottodriver @swanilla are you guys using recaptcha solvers?
I use an injected Chrome extension (Capsolver) as part of my browser instance, which seems to operate fine for the v2 reCAPTCHA, but it does not take care of the v3.
Same issue here -- the --user-data-dir
argument used to result in better Recaptcha scores earlier, however lately that's also not helping.
Not using this library right now but note that: Seems to be working using same country and bordering countries proxies, anything else got low score / challenge triggered
have anyone try setting turnstile to false?
have anyone try setting turnstile to false?
I've tried numerous configurations, probably tried this at some point as well, but nothing helps.
Description
I am running a web app that launches the puppeteer-real-browser instance on demand, i.e. when an API endpoint is hit. The puppeteer should then sign in to a web app that uses CAPTCHAs to protect their sign in page.
I use an extension to solve v2 CAPTCHAs (Capsolver), which does a good job most of the time. But even when it solves it, and I am able to click on the "Log in" button, I get back an error from the web app's API, something like "captcha incorrect"
I suspect it has something to do with the browser's reCAPTCHA v3 score being very low - almost always 0.1 when I check.
What can I do to improve this?
This seems valid for both Docker and local executions.
Full steps to reproduce the issue
Here is how I launch the browser:
Issue Type
Performance, Support
Operating System
Linux, Mac OS
Do you use Docker?
I don't use Docker