ResearchExplainer
Epoch: GLM 5.3 turned Firefox patches into exploits for $1,000
The open-weight model built working attacks on six Firefox releases from public fixes before users received them, in up to 48 hours per attempt.
Z.ai
Every ClaudeGPT Si story about Z.ai, newest first: news, analysis, reviews and guides.
3 stories · Updated
GuidesGuide
GLM 5.3 how to run locally: 753B weights, 8 GPUs, vLLM or SGLang
The model cards give no hardware list. The vLLM and SGLang recipes start at eight H200 GPUs for GLM 5.3, and Flash is smaller but still a server job.
AI ModelsExplainer
Is GLM 5.3 open source? Weights, licenses, size and pricing
Z.ai put GLM-5.3 and GLM-5.3-Flash on Hugging Face. Flash is MIT, the full model has a custom license, and API prices start at $0.15 per million input tokens.