Databricks says GLM-5.2 matched Claude Opus 4.8 at 34% lower cost

The open-weight model from Zhipu AI scored 62.1% on SWE-bench Pro and 81.0% on Terminal-Bench 2.1, while Databricks reported up to 392 tokens per second in inference.

Summary

verifying reliability

Terms & Concepts

No specialized terms available for this topic.