As an effective alternative to fossil-fueled power plants, renewable energy generation, particularly photovoltaic (PV) power, has been widely promoted across different regions of the world. However, when participating in electricity markets, PV power plants face various uncertainties, including their production levels, balancing market demand, market prices, and competitors’ bidding behaviors. Existing research on strategic bidding has largely overlooked the proactive decision-making of competing generators, making it difficult to accurately reflect the strategic bidding behavior of market participants. To address this issue, this paper proposes a two-stage Long Short-Term Memory (LSTM)-based method for predicting the bidding behavior of equivalent competitors. In the first stage, a bidding estimation model is developed to estimate the bidding function of the equivalent competitor (EC) unit from the previous hour, ensuring that the equivalent market closely resembles the actual market. In the second stage, to reveal the relationship between net load and competitor bidding, the estimated equivalent market data from the first stage is utilized to model the behavior of EC units and predict their bidding strategy for the next hour. Case study results demonstrate that the proposed two-stage LSTM model achieves high accuracy in modeling the bidding behavior of equivalent competitors.