Threads in Windows run until either their quanta (time slice) ends, they block (e.g., doing i/o that hasn't completed) or they're interrupted to run something higher priority that's just become ready.
Windows allows bumping the priority of a thread, all the way up to THREAD_PRIORITY_TIME_CRITICAL. But even the highest priority threads are occasionally interrupted to run lower-priority threads by the Windows scheduler, which uses random boosts to avoid a deadlock condition called priority inversion.
How or when an application creates new threads and what they do is a design decision embedded into the internal logic of the program, not something you can control except perhaps through the way you use the application, e.g., opening more tabs or whatever.
Bottom line is that if you're wondering what it takes to max out CPU usage as much as possible, the answer is takes a CPU-intensive activity with as many threads as you have processors, doesn't block on i/o and runs at higher priority (e.g., just being the foreground app) than other tasks.