Z.ai: GLM 5

Z.ai's GLM-5, released on February 11, 2026, is positioned as a flagship open-source foundation model for complex systems design and long-horizon agent workflows, targeting expert developers. With a substantial context length of 204800, it demonstrates exceptional reliability, achieving a 97% success rate across benchmarks, indicating consistent operational stability. However, its speed performance is a notable weakness, ranking in the 7th percentile, suggesting longer response times compared to most models. Price data is currently unavailable, hinting at potential free-tier usage. GLM-5 excels in several critical areas. It achieves perfect accuracy in Ethics and strong performance in Reasoning (89th percentile), even ranking in the top 3 for cost efficiency in the latter. Its Coding capabilities are also robust, scoring 92.0% accuracy (79th percentile). While its Instruction Following is good at 75.8% (85th percentile), its General Knowledge (97.5% accuracy) and Email Classification (97.0% accuracy) are solid but fall closer to the median. A key area for improvement is its handling of hallucinations, where it scored 84.0% accuracy, placing it in the lower 31st percentile, indicating a tendency to provide information for fictional concepts rather than acknowledging uncertainty.