This paper proposes an approach to categorize electricity suppliers (ESs) for energy trading between ESs and a single aggregator. A principal-agents game model is developed to model the interactions between an aggregator and different categories of ESs by considering the benefits of both parties. In a proposed game, the aggregator as a principal will purchase a certain amount of power from different-category ESs with the cheapest pricing options available, and at the same time the ESs, acting as agents will maximize their utilities by selling their power to the aggregator instead of feeding the grid at a low rate. The developed optimal contract-based scheme, which can be implemented distributed manner, allows different-category ESs to sell their power at different prices based on their unit production cost to maximize their benefits, and the total cost to the aggregator is minimized. Numerical analysis confirms the effectiveness of the proposed ESs categorizing framework in the development of a contract-based incentive mechanism for energy trading.