Databricks has selected the Chinese open-source model GLM 5.2 as its default coding engine after it demonstrated performance on par with Anthropic’s Opus 4.8 at a significantly lower cost. According to The Decoder, Databricks benchmarked these coding agents using its own multi-million-line codebase, finding that GLM 5.2 completed tasks at $1.28 each compared to Opus 4.8's $1.94.
The Decoder also reported that Databricks plans to deploy GLM 5.2 as a daily coding workhorse, signaling a strategic shift towards more cost-efficient AI tools in software development. This move highlights the growing impact of Chinese AI models in global tech infrastructure.
For Japanese markets, this development underscores the increasing competition among AI coding models, which could influence cost structures and innovation pace in Japan’s tech and financial sectors, particularly in areas like algorithmic trading and fintech development.