G protein-coupled receptors (GPCRs) are crucial players in various physiological processes, making them attractive candidates for drug discovery. However, traditional approaches to GPCR ligand discovery are time-consuming and resource-intensive. The emergence of artificial intelligence (AI) methods has revolutionized the field of GPCR ligand discovery and has provided valuable tools for accelerating the identification and optimization of GPCR ligands. In this study, we provide guidelines for effectively utilizing AI methods for GPCR ligand discovery, including data collation and representation, model selection, and specific applications. First, the online resources that are instrumental in GPCR ligand discovery were summarized, including databases and repositories that contain valuable GPCR-related information and ligand data. Next, GPCR and ligand representation schemes that can convert data into computer-readable formats were introduced. Subsequently, the key applications of AI methods in the different stages of GPCR drug discovery were discussed, ranging from GPCR function prediction to ligand design and agonist identification. Furthermore, an AI-driven multi-omics integration strategy for GPCR ligand discovery that combines information from various omics disciplines was proposed. Finally, the challenges and future directions of the application of AI in GPCR research were deliberated. In conclusion, continued advancements in AI techniques coupled with interdisciplinary collaborations will offer great potential for improving the efficiency of GPCR ligand discovery.