Zhipu's Latest AI Model Breaks Performance Barriers

In a significant milestone for China's AI industry, Zhipu AI has released its GLM-5.1 large language model that now leads global rankings for professional software development capabilities. The model achieved top scores on the rigorous SWE-bench Pro benchmark, outperforming previous leader Claude4.6Opus from Anthropic.

Pricing Power Returns to Chinese AI

The technological breakthrough comes with an interesting market development - Zhipu has increased prices by 10% across its GLM series models. This adjustment brings their pricing for core coding applications nearly equal to Anthropic's Claude3.5Sonnet.

"This marks a turning point," observed industry analyst Li Wei. "Chinese firms are no longer competing solely on price - they're proving they can command premium pricing through superior performance."

Engineering Without Limits

What sets GLM-5.1 apart is its remarkable endurance and autonomy:

• 8-hour continuous operation on single tasks
• Self-directed planning and execution
• Error self-correction capabilities
• Strategy adaptation during long workflows

Unlike typical AI models designed for short interactions, GLM-5.1 can manage complex engineering projects from start to finish, delivering complete solutions rather than just suggestions.

Market Responds Enthusiastically

The announcement sent Zhipu's stock soaring 14.12% on August 8th, closing at ¥890 per share and pushing the company's market valuation to ¥396.4 billion. Just one year after slashing prices by 90% to gain market share, Zhipu has successfully transitioned to competing on technological merit.

Key Points:

• 🏆 GLM-5.1 tops SWE-bench Pro rankings globally
• 💰 10% price increase aligns with international competitors
• ⚙️ Unprecedented 8-hour autonomous operation capability
• 📈 Stock jumps 14% as investors reward performance focus