Open access
Sep 2026
Applications of large language models in anxiety and depression patient care: a cross-model comparative analysis of dialogue quality
GPT-5.3-Codex and Claude-Opus-4.6 demonstrated the highest comprehensive efficacy under the standardized simulated consultation framework, which may inform future model optimization and prospective validation.
Jie Min, Rong-Rong Jiang, Tao Yang et al.
· Scientific Reports · 0 citations