Nobody sells asmack (as far as I know). I think you may be misunderstanding some things.
Smack is an XMPP Library implemented in Java developed here by the igniterealtime.org community and currently led by Flow and was led for previously by rcollier for a long time. Smack is the name of this library, and the word “Smack” does not apply to any other XMPP library (there are many XMPP libraries).
Openfire is an XMPP server built using the Smack library. Spark is an XMPP client built using the Smack library. There are many other programs in the wild that use Smack as their core for XMPP.
ASmack is the Android version of Smack, developed by Flow. (Flow correct me if I"m wrong), I believe ASmack was a fork of Smack to make it compatible on Android devices. I believe currently Flow is working to merge the two back into a single library that can work on Android as well as regular devices/machines.
These are all open source software/libraries.
Many vendors, such as QuickBlox and Cisco, take these open source software/libraries and use them in their paid products, sometimes with minor modifications, sometimes without any modifications at all.
In the case of QuickBlox, it appears that they have taken Smack (the XMPP library developed here) and have used that in their product. They also may be using Openfire behind the scenes as their server. What you get when you pay them for their service, is an abstracted API they have bolted on-top of Smack/Openfire. They do this to better control what you can or can not do with their service. Essentially, you are paying them for the same underlying products (Smack and Openfire).
So the “better documentation” that QuickBlox has provided, is really only the documentation they have made for their abstracted API. Under the covers, your API calls are going to be translated into Smack API calls.
So, if you feel more comfortable paying someone to run the server and provide you with a different API, well, that may work for you. Just know, that in doing so, you may make it much more difficult to switch away from using their service in the future (their API is non-standard and non-portable to any other XMPP service because it’s their own custom API).
To address a point in another post you made, developing a full featured instant messanging client is going to require some time. You will need to first understand XMPP, and either write your own XMPP library, or use one that exists already (like Smack). Then on top of that you will build how it works and functions, the UI, etc. It seems like you have a tight deadline to have a working XMPP client. I don’t think this is something you can just “hack” together overnight. It may take weeks or months to develop a full featured client that does everything you want. Spark has been developed for years and still has bugs (like all software). Depending on what your project is, you may be better off taking an existing XMPP client (like Spark) and then modifying it until it suits your needs. But, this will require you then learn how Spark operates. There is no getting past the steep learning curve when you jump into a new codebase.