since I already have plaintext and ciphertext why would I need the key for? what purpose would that serve?
The intuition behind known-plaintext attacks is that often an attacker is aware of the plaintexts of some observed ciphertexts (generalizing to being able to /choose/ those, for chosen-plaintext attacks), but might also have observed ciphertexts for which they do not know the corresponding plaintext. Consider for example an online document editor storing users' documents in an encrypted form. You as an attacker can easily get that service to encrypt whatever plaintext you want by having it store a document of your choice. But you do not yet have any information about other users' documents.
If now a known-plaintext (or chosen-plaintext) attack allows you to e.g. derive the key, you can use this knowledge to also decrypt all other ciphertexts for which that key had been used.
See e.g. this post and its answers for additional insight into how known-plaintext attacks map to "real" attack scenarios.