Introducing GLM-4.6: Advanced Agentic, Reasoning and Coding Capabilities As our new flagship model, GLM-4.6 brings significant advancements across real-world coding, long-context processing (up to 200K tokens), reasoning, search, writing, and agentic applications. API:

To better handle user feedback, we’ve set up a small team to quickly process unresolved email inquiries. If you have an urgent issue that hasn’t been addressed (reported on X, LinkedIn, Discord, or elsewhere), you can now email user_feedback@z.ai. We’ll provide an initial

The MTP layers for GLM-4.6-FP8 have been successfully built and deployed to the repository. https://t.co/M94rx9Tim0

GLM-4.6 Global Fan Giveaway! 🎉 https://t.co/rXVh6HmBJ6 Limited-Edition Panda Keychain — exclusively for GLM-4.6 fans! Come meet us in person at the GITEX booth and grab yours! 🐼✨

GLM Coding Plan now supported👇

A Chinese AI model just shattered our assumptions about the AI market. GLM-4.6 recorded the fastest adoption curve we've seen: 168 million → 15.9 billion tokens in 12 days. That's 94x growth. Here's why every developer needs to pay attention 🧵

On webdev arena, GLM is ranked #4 and #1 in all open models

The World’s Top-Ranked AI Coding Solution is coming to Dubai! Meet https://t.co/IQMnfBczAT at #GITEXGLOBAL 2025 - Booth CC1-89 Explore: Sovereign AI • AI Coding • Industry Agents 📍 Dubai World Trade Centre | Oct 13–17 📧 enterprise@z.ai #SovereignAI #AICoding

GLM-4.6 performs pretty well on Design Arena @designarena_ai. Big thanks to everyone supporting open models!

Let’s co-design the future. What’s on your mind?

We are developing the next version of GLM. What features do you want the most?

Starting today, you can use any open-source model to power your Droids. Droids achieve the highest scores across all open-source models on Terminal-Bench. We find GLM 4.6 to be the most performant, remarkably achieving a score in Droid that beats Sonnet 4 in Claude Code.

🚨 New Top Open Model Update! A relative newcomer to the Arena, @zai_org's GLM-4.6 takes the clear, undisputed #1 spot for Top Open Model. 🏆 It also ranks #4 overall, which is not an easy feat! The next top open model, DeepSeek R1 0528, has been the standing champion for

SORA + GLM-4.6 is next-level 🔥 This duo is insane for producing cinematic, high-quality videos no studios, no teams, no $1K/hour directors. Just two AI tools working together like a full creative department. Here’s how it works: → GLM-4.6 expands your concept into full

We hit #1 on @huggingface again, and it's all thanks to our community's feedback and contributions❤️ https://t.co/xsf4ia6fny

Service in https://t.co/gocggreDem is back

We're experiencing an attack on our CPU servers, causing a temporary outage for https://t.co/IQMnfBczAT Chat. We're working on a fix and will be back ASAP. Thanks for your support and understanding.

GLM-4.6 is ranked No.4 at LLM arena ( https://t.co/kvGWhr6snF), and No.2 when style control removed. Welcome to give it a try at https://t.co/QaycfjBGqO

so we analyzed millions of diff edits from cline users and apparently GLM-4.6 hits 94.9% success rate vs claude 4.5's 96.2%. to be clear, diff edits are not the end-all-be-all metric for coding agents. but what's interesting is three months ago this gap was 5-10 points. open

GLM-4.6 实测! 最近都很流行更新一大截然后版本就增加0.1嘛? 直接说结论, 前端能力非常好, 效果甚至达到了炫技的程度, 不过随之而来的是复杂场景可能面临性能问题, 所以需要注意防止模型过度发挥或者偶尔留意下代码设计和架构. 不过 python 能力没看到太多变化, 考虑到 python 就一个测试,

GLM-4.6 is now live in Kilo Code, and the economics are wild: 48.6% win rate against Claude Sonnet 4 $0.60/$2.20 per million tokens (vs Claude's $3/$15) 30% more token efficient 200K context window This changes the math on AI-assisted development. 🧵