The present study compared individual and collaborative vocabulary learning through two types of reading-based pedagogical tasks in an online elementary Arabic classroom. Fifty-two beginner learners of Arabic completed a pretest, a story-based task (either multiple-choice or gap-fill) with embedded novel vocabulary, and two posttests over a 2-week period. Tasks were performed either collaboratively in groups (n = 32) or individually (n = 20) during class time. While there was no effect of task type, results showed that the collaborative condition was more conducive to vocabulary gains than the individual condition. From a processing perspective, two vocabulary frameworks were compared: The "Involvement Load Hypothesis" (ILH) and the "Technique Feature Analysis" (TFA). Vocabulary gains were found to be more clearly predicted by the TFA as it included components that captured the cognitive loads of collaborative tasks. The study provides implications for teaching Arabic vocabulary, particularly with novice learners.