Chinese AI lab Z.ai released GLM-5.3, a 743-billion-parameter coding model it claims is the most capable open-weights coder on the market. The model scores 34.5% on the Z.ai Code Bench at Max effort, up from 23.4% for its predecessor GLM-5.2, while consuming fewer tokens per task (75,000 versus 96,000). In cybersecurity, GLM-5.3 detected 2,436 vulnerabilities across 269 open-source projects, including 1,097 of medium-to-high severity. The model is available now through GLM Coding Plan and ZCode, with open weights set for public release in about two weeks after a safety review. API pricing is estimated at roughly one-tenth of U.S. frontier model rates.
Source: Read the original article
