GLM 5.1

• 🆕 Z.ai's next-generation flagship for agentic engineering and long-horizon coding.
• 🧠 Mixture-of-experts design carried over from GLM 5.
• 📏 Roughly 200K-token context window with FP8 quantization.
• 🔧 Built-in function calling, tool use, and web search/browsing.
• 🎯 Vendor reports a SWE-Bench Pro score of 58.4 (self-reported).
• ⚡ Designed to sustain optimization across many reasoning rounds and tool calls.
• 🔒 Released under the permissive MIT open-weight license.
• 📚 Tuned for multi-stage software tasks and math reasoning.

Z.ai

Z.ai, formally Knowledge Atlas Technology Joint Stock Co., Ltd., is a Chinese technology company specializing in artificial intelligence. Previously known internationally as Zhipu AI, the company rebranded to Z.ai in 2025. Its core focus is the GLM family of…

GLM 5.1 is the latest entry in Z.ai's GLM series, positioned as a flagship model for agentic engineering, complex coding, and long-horizon reasoning. It is an incremental update to its same-family predecessor GLM 5, retaining the same underlying mixture-of-experts approach described on the family's model cards. The model is distributed in an FP8 quantized form and exposed with a context window of roughly 200K tokens.

According to Z.ai, GLM 5.1 offers stronger coding capabilities than GLM 5, with better judgment on ambiguous problems and the ability to break tasks down, run experiments, read results, and revise strategy across extended sessions. The provider describes it as designed to sustain optimization over many reasoning rounds and tool calls, becoming more effective the longer it runs.

On vendor-reported benchmarks, Z.ai cites a SWE-Bench Pro score of 58.4 and, on KernelBench Level 3, a 3.6× geometric-mean speedup driven by thousands of tool-invocation-based optimizations. Note that these are the lab's own figures rather than independent measurements.

Beyond GLM 5, the broader family includes earlier releases such as GLM 4.7. Like its siblings, GLM 5.1 ships under the MIT license, permitting commercial use, modification, and redistribution.