Zixuan Li says GLM-5.3 has found 4,249 potential vulnerabilities in 389 open-source projects

ZZixuan Li9/29/2026, 19:21:29

Zixuan Li says GLM-5.3 has helped defend 389 open-source projects, with 4,249 potential vulnerabilities found so far.

OpenVuln is still running. The service remains free, and findings go privately to maintainers. It is available as the OpenVuln Hugging Face Space.

#glm-5.3 #openvuln #open-source #security

View original post ↗

Previous articleArtificial Analysis open-sources AA-AgentPerf-Local for laptop and workstation inference Next articleArtificial Analysis: Intelligence Index high rose from 17 to 58 in a year

Related AI news

GLM Coding Plan bills peak-hour usage at non-peak rates until Oct. 7

Zixuan Li says peak-hour usage on GLM Coding Plan is billed at non-peak rates until Oct. 7. All users have also received four weekly reset cards and four five-hour reset cards.

ZZixuan Li•9/28/2026, 07:38:54•

#glm #coding-plan #pricing

GLM Coding Plan users receive four weekly and four five-hour reset cards

Zixuan Li says every GLM Coding Plan user has received four weekly reset cards and four five-hour reset cards, usable in all supported agents. Users who return within the next month get the same benefits.

ZZixuan Li•9/28/2026, 06:27:33•

#glm #coding-plan #reset-cards #usage-limits

Faster GLM-5.3-Flash now live at up to 200 tokens/s as glm-5.3-flashx

Zixuan Li (@ZixuanLi_) said Faster GLM-5.3-Flash is now live, with model code glm-5.3-flashx and speeds up to 200 tokens/s. It is priced at 2.5× GLM-5.3-Flash on both the Coding Plan and API, open to all API users, while Coding Plan users can apply via a Google Form.

ZZixuan Li•9/21/2026, 09:12:36•

#glm #glm-5.3-flash #api #coding-plan

NVIDIA paper details VERA framework for co-evolving agent harnesses and models

An NVIDIA research paper presents VERA, a system that converts benchmark trajectories into over 9,000 restartable sandboxes with rubric scoring while co-evolving agent harnesses alongside model weights. The team open-sourced the environment corpus, reporting that a 27B agent scored 71.6 on AutoCoWorkBench to top Claude Opus 4.8.