On average, the Kaplan CATs seem to be harder than the actual GMAT, and most people score somewhat lower on Kaplan tests than on the GMATPrep or the actual tests. But you may or may not be "average."

It isn't strange at all for Kaplan tests to overstate your scores, depending on your exact strengths and weaknesses.
I think its always best to avoid taking your practice test scores too seriously, unless you're doing the GMATPrep. I would argue that the
MGMAT tests tend to be more accurate than the Kaplan tests on average... but again, you might not be "average," and the
MGMAT tests might wildly understate or overstate your actual skills.
The GMAT is a really complicated test, and the exact scoring algorithm is a closely guarded secret. Tests from Kaplan or
MGMAT represent a sincere effort by very brilliant people to reverse-engineer the GMAT... but they aren't the actual test, and in some ways, they aren't even close.
I know that it's really hard to accept this advice, but try not to freak out about your Kaplan or
MGMAT practice test scores--they don't always mean exactly what they seem to mean. Analyze your errors carefully, take a critical look at your timing habits, and try to squeeze every bit of wisdom you possibly can out of your practice tests. But don't be too shocked if your Kaplan (or
MGMAT or PR or Veritas) test results don't ultimately match your GMATPrep or actual GMAT scores. They're great practice, but they're rarely going to offer a perfect indication of your GMAT skills.