Immunotherapy with immune checkpoint inhibitor (ICI) drugs is gradually becoming a hot topic in cancer treatment. To comprehensively evaluate the safety and efficacy of ICI drugs, we employed the Bayesian model and conducted a network meta-analysis in terms of progression-free survival (PFS), overall survival (OS) and severe adverse events (AEs). Our study found that treatment with ipilimumab was significantly worse than standard therapies in terms of PFS, whereas treatment with cemiplimab significantly improved PFS. The results also indicated that cemiplimab was the best choice for PFS. Treatment with nivolumab, pembrolizumab and nivolumab plus ipilimumab significantly improved OS compared to standard therapies. In terms of OS, cemiplimab was found to be the best choice, whereas avelumab was the worst. In terms of severe AEs, atezolizumab, avelumab, durvalumab, nivolumab, and pembrolizumab all significantly reduced the risk of grade 3 or higher AEs compared to standard therapy. The least likely to be associated with severe AEs were as follows: cemiplimab, avelumab, nivolumab, atezolizumab, and camrelizumab, with nivolumab plus ipilimumab to be the worst. Therefore, different ICI drug therapies may pose different risks in terms of PFS, OS and severe AEs. Our study may provide new insights and strategies for the clinical practice of ICI drugs.