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-5.1
Z.ai • April 2026
Training Data
Up to early 2026
GLM-5.1
April 2026
Parameters
754 billion (40B active)
Training Method
MoE with post-training optimization on Huawei Ascend 910B chips
Context Window
200,000 tokens
Knowledge Cutoff
Not disclosed
Key Features
#1 SWE-Bench Pro (58.4) • 8-Hour Autonomous Execution • 1,700 Sequential Agentic Steps
Capabilities
Coding: Outstanding
Reasoning: Excellent
Agentic Tasks: Excellent
What's New in This Version
Achieves 28% coding improvement through post-training alone, with stronger multi-file refactoring and ability to sustain 1,700 autonomous steps over 8 hours vs GLM-5's ~20
Zhipu AI's current flagship refined from GLM-5 for agentic engineering, topping the SWE-Bench Pro leaderboard with sustained 8-hour autonomous execution
What's New in This Version
Achieves 28% coding improvement through post-training alone, with stronger multi-file refactoring and ability to sustain 1,700 autonomous steps over 8 hours vs GLM-5's ~20
Technical Specifications
Key Features
Capabilities
Other Z.ai Models
Explore more models from Z.ai
GLM-5.3
Z.ai's flagship on GLM-5.2's base, with every gain from scaled-up post-training — frontier-level coding and cyber-defense capability
GLM-5.3-Flash
Z.ai's MIT-licensed efficient model — natively multimodal, with hybrid sparse and linear attention far cheaper to run than GLM-5.3
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
