What is the difference between ACA and CBC in the term of accuracy? I’ve read the manual and noticed that each had totally different features. But the problem is that I didn’t get the exact point of them. My survey will have 4 or 5 attributes, with 4 or 5 levels per attribute, and approximately 100 respondents. 

You can read about some of the advantages and disadvantages of the different methods of conjoint analysis in our paper: Which Conjoint Method Should I Use? (2013)

Adaptive Conjoint Analysis (ACA) and Choice-Based Conjoint (CBC) are both perfectly capable of getting accurate estimates of respondents’ part-worth utilities. They are simply different methods of getting people’s preferences. ACA is an adaptive, ratings-based approach that will show people partial profile cards for ratings based tradeoffs. CBC is a full profile choice based method that will ask people to select their favorite concept in a set of concepts.

If pricing research is a big part of your studies then I would definitely favor CBC over ACA. With only five attributes you should be able to do a CBC study, even with only 100 respondents. If you are concerned about sample size then ACBC does better with small samples then CBC.