Zhipu AI has released its flagship GLM-5.1 model, which reportedly outperforms Anthropic’s Opus 4.6 on the SWE-bench Pro software engineering benchmark.The company also raised pricing by 10%, with coding token costs now approaching levels of Anthropic’s Claude Sonnet 4.6, according to OpenRouter.(金十数据)