The only code I had was the math for the inverse kinematics. Which actually is not that hard, there are lots of websites that explain how it works.
My biggest problem was coding the walking gait. I first wrote a sketch using a for loop and the delay function to move the servos. That was pretty easy. But when you want the micro controller to do a bunch of other things you're in trouble. So I tried dozens of ways to code the gait using the millis timer, after a dozen revisions of my code I finally got it right.
I did look at the Lynxmotion code before I built my hexapod, but it's pretty complex. I can't understand much of it.
I name it a monster because it uses so much current. I once tried to power it without the UBECs, and connected the servos directly to my lab power supply. I saw the current shoot up to almost 20A. I'm using very cheap metal geared servos (standard size), Towardpro MG996r servos. They cost less then 5€ a piece and have +10kg/cm torque.