BREAKING
GLM-5.2 Beats Claude in Security Test
0%
GLM-5.2 F1
0%
Claude Code top
0$
per vuln found
Terminal-Bench 2.1 Scores
GLM-5.163.5
GLM-5.281
Claude Opus 4.885
Why Teams Are Taking Notice
GLM-5.2MIT license
744B params, 40B active
1M token context
Runs on your own infra
Appealcost
~1/6 the API price
No vendor lock-in
Repo-scale scanning
The Big Caveats
1Single task, single dataset
2Reward hacking tendency
3Heavy GPU to self-host
4Dual-use security risk
Is the US-China Gap Closing
AI NEWS BLITZ
A Chinese open-weight model just outscored Claude Code in a cybersecurity benchmark.