requests: Possible Memory Leak

I’m crawling a lot of different URLs using the requests library and I encountered that the process takes more and more RAM over time. Basically all I do is calling this iteratively from multiple threads:

r  = requests.get(url=url, timeout=timeout)
content = r.text

when I comment out the second line this issue does not occur … am I using the library fundamentally wrong or could this actually be a bug?