Back
RCreddit.com
21
·9 hr ago·Dev community · RSS

zai-org/GLM-5.3 · Hugging Face

View original
Hugging FaceModel releaseOpen source

Heat trend

Collecting trend data

The percentage is based on available heat signal, not comment count or independent people.

Why it matters

Hugging Face model activity is surfacing — worth tracking for capability changes, ecosystem impact, and availability.

AI summary

GLM-5.3, developed by zai-org, is a new open-weights model that significantly improves upon GLM-5.2, particularly in complex coding and long-horizon tasks. It utilizes the same base model as GLM-5.2, with all enhancements stemming from post-training. GLM-5.3 demonstrates a 50% improvement over GLM-5.2 on the Z.ai Code Bench and achieves open-source SOTA on public benchmarks like Terminal Bench 3.0 and Agents' Last Exam, making it the most capable open-weights model for coding.

GLM-5.3 uses the same base model as GLM-5.2 — every gain comes from post-training. Compared with GLM-5.2, it is much better at complex coding and long-horizon tasks:

- Stronger Coding: GLM-5.3 is the most capable open-weights model for coding, with a 50% improvement over GLM-5.2 on our in-house Z.ai Code Bench. It also achieve open-source SOTA on public benchmarks including Terminal Bench 3.0 and Agents' Last Exam.

- Emergent Cyber Capability: As we scaled post-training, cyber capability developed faster than we expected. GLM-5.3 is state of the art on CyberGym for vulnerability discovery, and its gains are largest further up the exploitation chain, where it more than doubles GLM-5.2 on exploitation benchmarks.

zai-org/GLM-5.3 · Hugging Face · BuzzRadr