Zhipu AI Releases Open-Source GLM-5 MoE Model
The new Mixture-of-Experts model from the Chinese AI company combines an advanced architecture with a fully permissive MIT license for commercial use.
Chinese AI research firm Zhipu AI has released GLM-5, a significant new open-source language model. This release marks a major addition to the open-source ecosystem, offering a sophisticated new architecture from a prominent AI company.
Efficient by Design
GLM-5 is built on a Mixture-of-Experts (MoE) architecture, a technique that improves computational efficiency by activating only specialized parts of the network for any given task. The model also incorporates sparse attention, further optimizing its performance and resource usage. This combination is designed to deliver powerful reasoning capabilities while maintaining manageable inference costs.
Perhaps most notably, Zhipu AI has released GLM-5 under the permissive MIT license, which allows for broad and unrestricted commercial application. The choice of such an open license for an advanced model encourages wide adoption and makes it a valuable asset for developers and businesses building with AI.
The release of GLM-5 provides a new, highly capable foundation for the open-source community, blending an efficient, modern architecture with the freedom to innovate and build commercial products.
Sources
- Visit
zai-org/GLM-5
Hugging Face
More in Text / LLM
Meituan Ships a Lighter, Sparser LongCat-Flash
The food-delivery giant's newest open model trims its mixture-of-experts design for more efficient inference under an MIT license.
DeepSeek Refreshes V4-Flash With New 0731 Checkpoint
The MIT-licensed mixture-of-experts model returns in an updated build shipping with FP8 weights for cheaper inference.
DeepSeek Ships V4-Flash, a 304B MoE Tuned for Agents
The latest checkpoint in DeepSeek's V4 line leans into agentic workflows while keeping the permissive MIT license.
0 comments
No comments yet. Be the first to weigh in.