Self-Reference Count
Self Reference
Pass rate
0%
Survived 10 out of 15 breakers
GLM-5 is Z.ai’s flagship open-source foundation model engineered for complex systems design and long-horizon agent workflows. Built for expert developers, it delivers production-grade performance on large-scale programming tasks, rivaling leading closed-source models. With advanced agentic planning, deep backend reasoning, and iterative self-correction, GLM-5 moves beyond code generation to full-system construction and autonomous execution.
202,752 tokens
$0.80 /1M tokens
$2.56 /1M tokens
–
| Test | Category | Latest Result | Success Rate | |
|---|---|---|---|---|
| Self-Reference Count | Self Reference | 0% | ||
| Bullshit Detector | Epistemic Humility | 0% | ||
| Silence Protocol | Instruction Following | 13% | ||
| Contradictory Premises | Logic Reasoning | 13% | ||
| 10-Step Instructions | Instruction Following | 25% | ||
| The Missing A | Pattern Matching | 50% | ||
| Car Wash Dilemma | Logic Reasoning | 75% | ||
| Strawberry Problem | Character Counting | 100% | ||
| Reverse Word Test | Character Manipulation | 100% | ||
| Alice's Brother Problem | Logic Reasoning | 100% | ||
| Broken Mug | Lateral Thinking | 100% | ||
| Horse Race Logic | Logic Reasoning | 100% | ||
| The Compartment Trick | Logic Reasoning | 100% | ||
| Sycophancy Trap | Logic Reasoning | 100% | ||
| Coin Flip Paradox | Logic Reasoning | 100% |