GLM 5 by Z.AI (Zhipu AI) is a 744B MoE foundation model for agentic engineering, with DSA sparse attention, async RL via Slime framework, and 200K context.
Capabilities, design details, and architectural traits
GLM 5 is Z.AI's (Zhipu AI) next-generation foundation model explicitly designed to transition the paradigm of "vibe coding" to agentic engineering — shifting from passive code generation toward autonomous, long-horizon software engineering execution.
GLM 5 builds on the ARC (Agentic, Reasoning, and Coding) capabilities of its predecessor. The Slime framework enables the model to continuously improve through extended interaction chains — not isolated turn-by-turn feedback. The model supports switchable thinking modes (enabled / disabled) per request via the API.
On SWE-bench Verified, GLM 5 scores 77.8, and on Terminal Bench 2.0, it scores 56.2 — highest among open-weight models at release and surpassing Gemini 3.0 Pro overall. On BrowseComp, MCP-Atlas, and τ²-Bench, it ranks first among open-weight models. Real-world software engineering usability is documented as approaching Claude Opus 4.5.
Independent evaluations · Artificial Analysis
Evaluate specifications, pricing, and independent benchmark indices
| Model Details | |||
|---|---|---|---|
| General Info | |||
| Provider | Z AI | Anthropic | Anthropic |
| Release Date | February 11, 2026 | July 24, 2026 | June 9, 2026 |
| Knowledge Cutoff | - | May 2026 | - |
| Context & Limits | |||
| Context Window | 203K | 1M Best Context Window | 1M Best Context Window |
| Pricing (per 1M tokens) | |||
| Input Pricing | $1 Best Input Pricing | $5 | $10 |
| Output Pricing | $3.20 Best Output Pricing | $25 | $50 |
| Modalities | |||
| Inputs | text | textimage | textimagefile |
| Outputs | text | text | text |
| Benchmarks (0-100) | |||
| Intelligence Index | 33.2 | 63.1 Best Intelligence Index | 62.1 |
| Coding Index | - | 78.0 Best Coding Index | 76.5 |
| Agentic Index | - | 59.2 Best Agentic Index | 56.6 |
Graduate-level reasoning and expert Q&A evaluation.
Extremely difficult logical reasoning and knowledge.
Logical reasoning over long context windows.
Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.
