CharacterGLM Pro is Lingxin Intelligence's next-generation ultra-human model, officially launched after 6 months of product testing and optimization. This model leads in aspects such as coherence, human-like qualities, fluency, and engagement, achieving state-of-the-art (SOTA) levels. In cross-comparison tests, CharacterGLM Pro performed exceptionally well on consistency test sets, overall test sets, attack test sets, and online dialogue test sets, achieving the best levels across various metrics. The model utilized multiple test sets, including artificially generated data.
