Modulatory effects of transcranial magnetic stimulation (TMS) strongly depend on the stimulation parameters. Here, we compared the immediate, task-locked inhibitory effects on speech-related muscles and the tolerability of different TMS protocols during a language production task. Repetitive TMS (rTMS) and paired-pulse TMS (PP) were applied in 13 healthy subjects over the primary motor cortex (M1) during a finger-tapping/tongue-twisting tasks. The lowest subject-specific TMS intensity leading to movement disruptions was used for TMS over left-sided speech-related areas during picture naming. Here, time-locked PP and rTMS (10/30/50 Hz; randomized sequence) were applied. Cortical silent periods (cSPs) were analyzed from electromyography obtained from various face muscles. 30 Hz- and 50 Hz-rTMS reliably evoked tongue movement disruption (ICC = 0.65) at lower rTMS intensities compared to 10 Hz-rTMS or PP. CSPs were elicited from the left hemisphere by all TMS protocols, most reliably by PP (p < 0.001). Also, cSPs with longest durations were induced by PP. Exploratory analyses of PP suggest that the trials with strongest motor inhibitory effects (presence of cSP) were associated with more articulatory naming errors, hence hinting at the utility of TMS-elicited, facial cSP for mapping of language production areas. Higher-frequency rTMS and PP evoked stronger inhibitory effects as compared to 10 Hz-rTMS during a language task, thus enabling a probably more efficient and tolerable routine for language mapping. The spatial distribution of cranial muscle cSPs implies that TMS might affect not only M1, but also distant parts of the language network.