When preparing a movement, we often rely on partial or incomplete information, which can decrement task performance. In the behaving monkey we show that the degree of cued target information is reflected in both, neural variability in motor cortex and behavioral reaction times. We study the underlying mechanisms in a spiking motor-cortical attractor model. By introducing a novel and biologically realistic network topology where excitatory neuron clusters are locally balanced with inhibitory neuron clusters we robustly ensure multistable network activity across a wide range of network parameters. In application to the monkey task, the model performs target selection and accurately reproduces the task-epoch dependent reduction of trial-to-trial variability in vivo where the degree of reduction directly reflects the amount of available target information, while spiking irregularity remained constant throughout the task. In the context of incomplete cue information, the increased target selection time of the model explains the increased behavioral reaction time of the monkey. We conclude that context-dependent neural and behavioral variability are a signum of attractor computation in the motor cortex.