GLM 5.3 weights released publicly
Z.ai has released the weights for GLM 5.3, making the model publicly available. The release was announced on Hugging Face.
AI Lab
Official Z.ai announcements — model releases, product launches and research, each one summarized with every source covering it, by AIBriefs. RSS
Z.ai has released the weights for GLM 5.3, making the model publicly available. The release was announced on Hugging Face.
GLM-5.3, built on the GLM-5.2 base, shows a 50% improvement over GLM-5.2 on in-house coding benchmarks, making it the most capable open-weights model for coding.
Z.ai published GLM-5.3 weights on Hugging Face, its most capable model for agentic coding and cyber defense. Modular reports a 50% gain over GLM-5.2 on Code Bench; Together AI says it nearly matches Fable 5 at a fraction of the cost per task.
Z.ai's GLM-5.3 scores 60 on the Artificial Analysis Intelligence Index, up 7 points from GLM-5.2 and level with Kimi K3. It keeps GLM-5.2's 1M-token context and 128K max output while producing fewer output tokens at the same effort level.
GLM-5.3 scores 60 on the Artificial Analysis Intelligence Index, tying Kimi K3 and up 7 points from GLM-5.2, and hits 84.5% on the CyberGym vulnerability benchmark. It uses the same base model as GLM-5.2, with gains from post-training; weights release two weeks after launch.