Due to the poor quality of RGB images in complex scenes, leading to bad tracking performance. Therefore, the introduction of an infrared (TIR) modality complementary to RGB is crucial to overcome the limitations of single-modality tracking. To address this issue, we propose an RGB-T tracking method based on a dual-stream encoder structure, which aims to achieve multi-modality tracking by combining RGB and TIR. Its main contributions are as follows: Firstly, a dual-stream encoder structure based on an adapter fine-tuning strategy is proposed to reduce training cost while achieving cross-modality feature prompting. Secondly, a Modality Spatial Fusion (MSF) module is added after the backbone, which uses cross-attention mechanisms to enhance the learning ability of complex features and address tracking challenges in complex scenes. Our method was evaluated on the LasHeR and GTOT datasets, and a visual comparison with state-of-the-art algorithms was performed on the LasHeR dataset. The experimental results show that this method effectively addresses the performance limitations of single-modality tracking.