tech
Zhipu AI Unveils GLM-5.3-Flash, the Mystery Model Behind 'Ox Alpha'
Want to read in a language you're more familiar with?
TL;DR
- Zhipu AI's 'Ox Alpha' model is confirmed to be GLM-5.3-Flash.
- GLM-5.3-Flash is a mixture-of-experts model with 320 billion total parameters, 18 billion active per inference.
- It is optimized for coding, long agentic work, and a million-token context window.
- The model is natively multimodal, accepting text, images, video, and files.
- It features an architecture designed to reduce compute needs while maintaining capability.
- GLM-5.3-Flash shows strong benchmark performance in code and agentic tasks.
- Model weights are available on Hugging Face under the MIT license for local deployment.