In Internet of Things (IoT) environment there are a huge number of devices that need to communicate and send data continuously between these devices also between them and the cloud data center. This data is increasing exponentially, exposing the IoT environment to collapse. Therefore, so we need an environment that supports the stability of the IoT and at the same time increases the speed of exchange huge data between IoT devices. Fog computing and mist computing support real time data collection and analysis locally at IoT devices. In addition, fog computing and mist computing overcome on various challenges like network bandwidth, and reliability. Resource management and allocation for IoT tasks in three levels mist-fog-cloud architecture suffer from a lack of approaches and frameworks that handle this situation in an efficient manner. In order to address this shortage, Multi-Level IoT Tasks Scheduling (MLITS) model is introduced in this paper. MLITS is an orchestration system for managing and allocating IoT tasks over the mist-fog-cloud architecture. The proposed model performs IoT tasks based on their deadline and the urgency of their execution. In addition, it performs various types of IoT tasks without rapidly consuming available sources. Moreover, the proposed model maintains the resource usage balanced. Finally, The MLITS is simulated and evaluated on truthful fog resources and various workload circumstances. Also, the proposed scheduling model is compared with three scheduling model, namely; Min-Min, Credit-Based-Scheduling (CBS) and Earliest-Feasible-Deadline-First (EFDF). Through extensive simulations, we show that our proposed model enhances the performance metrics, namely; turnaround time, waiting time and throughput.