GLM-5.2
Zhipu AI's MIT-licensed flagship extending GLM-5.1 to a 1M-token context, the highest-ranked open-source model on long-horizon coding benchmarks
GLM-5.2
Z.ai • June 2026
Training Data
Not disclosed
GLM-5.2
June 2026
Parameters
753 billion (40B active)
Training Method
MoE with IndexShare sparse attention and improved MTP speculative decoding
Context Window
1,000,000 tokens
Knowledge Cutoff
Not disclosed
Key Features
1M-Token Context Window • IndexShare: 2.9x FLOPs Reduction • MIT-Licensed Open Weights
Capabilities
Coding: Outstanding
Reasoning: Outstanding
Agentic Tasks: Excellent
What's New in This Version
Quintuples GLM-5.1's context to 1M tokens while cutting per-token compute 2.9x via IndexShare, and beats GPT-5.5 on SWE-bench Pro (62.1 vs 58.6) at roughly one-sixth the cost
Zhipu AI's MIT-licensed flagship extending GLM-5.1 to a 1M-token context, the highest-ranked open-source model on long-horizon coding benchmarks
What's New in This Version
Quintuples GLM-5.1's context to 1M tokens while cutting per-token compute 2.9x via IndexShare, and beats GPT-5.5 on SWE-bench Pro (62.1 vs 58.6) at roughly one-sixth the cost
Technical Specifications
Key Features
Capabilities
Other Z.ai Models
Explore more models from Z.ai
GLM-5.1
Zhipu AI's current flagship refined from GLM-5 for agentic engineering, topping the SWE-Bench Pro leaderboard with sustained 8-hour autonomous execution
GLM-5
Zhipu AI's open-weight 744B MoE foundation model purpose-built for agentic engineering, trained entirely on Huawei Ascend chips
GLM-4.7
Z.ai's flagship model with industry-leading coding and multi-step task handling
