Semgrep finds GLM 5.2 outperforms Claude on security code analysis

Semgrep's internal benchmarks show Alibaba's GLM 5.2 model scoring higher than Claude on their cybersecurity-focused code tasks. The finding suggests open or cheaper model alternatives may be viable for security-focused development workflows, though results are specific to Semgrep's test suite.