Does the client need to install Cuda Toolkit to run the application?

I have finished building my cuda application and would like to install it on the clients PCs.
The cuda toolkit installer is about 2GB. It comes with the compiler, samples, tools, etc.
Is the whole toolkit definitly required to be installed on the client machine?
Is there no lighter version for just the CUDA Runtime API ?

A user does not have to have the CUDa toolkit installed to run a CUDA-accelerated application. Your CUDA-based application may rely on various dynamic link libraries associated with CUDA in which case you would need to re-distribute those. The CUDA documentation includes an overview of the re-distributable components in the EULA: