AI Model GLM-5.2 Beats Western Rival Opus 4.7 in Cost and Performance

Snowflake’s benchmark test pitted GLM-5.2 against Opus 4.7, a Western AI model. GLM-5.2 held its own, with 66% of tasks solved, despite being less consistent in its first attempts. However, GLM’s strengths lie in validating code across platforms, but its weaknesses include giving up too early and excessive checking. The Chinese model’s pricing puts pressure on Western AI companies, including Anthropic and OpenAI, which are facing serious pricing pressure and rely on revenue growth to justify their valuations.