Z.ai GLM

What hackers pick instead

• Qwen 3.8 27BAn open-weights model that runs on a single high-end GPU and nears frontier coding quality, though it overthinks, burns tokens and struggles with complex tasks.1173recommend
• KimiMoonshot's open-weight models rival top closed models at coding and research for a fraction of the price, though they can be slow and overthink.931recommend
• Gemma 4Google's open-weight model family runs well on local hardware with strong quality and token efficiency for its size, though critics cite unreliable tool calling and weaker coding than Qwen.713recommend
• QwenThe go-to open-weight model family for running capable coding and general AI locally on accessible hardware, though it falls short of frontier models like Claude Opus.474recommend
• Qwen3.6-35B-A3BA fast mixture-of-experts open model that runs well on consumer hardware and is surprisingly capable for its size, though it falls well short of frontier models on complex tasks.373recommend

Every quote

The 40 most detailed of 2560 comments. Each links to the original.

1. which was more like the ravings of a lunatic - but lo and behold, after much confusion and a dizzying number of eureka moments the task was completed very successfully

   Mixed · Jun 2026 · in “GLM-5.2 is the step change for open agents”

2. GLM 5.2 was used by Hugging Face for defense when they were being hacked by OpenAI because the guardrails on closed models meant they refused to help.

   Recommends · Sep 2026 · in “GLM-5.3 and the spread of advanced cyber capabilities”

3. other providers can offer third-party open source models with roughly the similar quality like this, Kimi K2.6 or GLM 5.1 for 10 times less the price

   Mixed · Apr 2026 · in “DeepSeek v4”

4. several Chinese models (at least Qwen, DeepSeek, Kimi and GLM) are "close enough" that with the right harness they are cost effective alternatives

   Recommends · Jun 2026 · in “Uber's $1,500/Month AI Limit Is a Useful Signal for AI Tool Pricing”

5. I've found GLM 5.2 to be better at security research than Opus 4.6 [4]. But it's a quirky model that degrades quickly at long context lengths.

   Mixed · Jul 2026 · in “UK AISI / Caisi Preliminary Assessment of Kimi K3's Cyber Capabilities”

6. I would constitute that even by using GLM 5.2 as you are originally doing even with API costs is probably much more sustainable over long run

   Recommends · Jun 2026 · in “Claude Code Is Steganographically Marking Requests”

7. I have been using a quorum composed of step-3.5-flash, Kimi k2.5 and glm-5 and I have found it outperforms opus-4.5 at a fraction of the cost

   Recommends · Feb 2026 · in “Nvidia and OpenAI abandon unfinished $100B deal in favour of $30B investment”

8. the fact that they've got an open model that is just a smidge less capable than the multibillion dollar state of the art models is fantastic

   Recommends · Dec 2025 · in “GLM-4.7: Advancing the Coding Capability”

9. Nothing felt off with GLM. It did what I wanted, was fast, had a decent not very annoying personality and was much cheaper than Opus or GPT.

   Recommends · Jun 2026 · in “Semgrep: GLM 5.2 beats Claude in our Cyber Benchmarks”

10. it was better than corporate's fave Sonnet 4.8 and a but worse than Opus 4.8. overall quite capable of a lot of what the dev team needed.

    Recommends · Jul 2026 · in “Open-weight AI is having its Kubernetes moment. Let's not ruin it”

11. The point isn't that the models are from China. It's that the weights are open. You can download the weights and finetune them yourself.

    Recommends · Feb 2026 · in “GLM5 Released on Z.ai Platform”

12. The best chinese models are deepseek (general purpose) and glm (coding) and they are both open weight and share lots of their tooling.

    Recommends · Jun 2026 · in “US Government directive to suspend access to Fable 5 and Mythos 5”

13. Not remotely usable interactively, but something I could plausibly push a codebase into and come back to a review in a couple of days.

    Mixed · Aug 2026 · in “GLM-5.3: Frontier Coding with Emergent Cyber Capabilities”

14. its one-shot performance is more impressive than its agentic abilities. On both metrics, GLM 5.1 is competitive with frontier models.

    Recommends · Apr 2026 · in “GLM-5.1: Towards Long-Horizon Tasks”

15. GLM 5.3 a lot trickier because if you don't want to have a model that is quantized too far down that is a fortune in VRAM and GPUs

    Mixed · Sep 2026 · in “GLM-5.3 and the spread of advanced cyber capabilities”

16. the obvious practical solution to the lack of frontier models in cyberdefense is that the defenders should run GLM 5.3 themselves

    Recommends · Sep 2026 · in “GLM-5.3 and the spread of advanced cyber capabilities”

17. GLM-5.2 meets my needs for "thinky" tasks, which for me is code and documentation reviews, technical chats and rubber ducking.

    Recommends · Jun 2026 · in “Anthropic updates their terms to verify age or identity”

18. Other than the top performers (e.g. GLM 5.1, Kimi K2.5, where required hardware is basically unaffordable for a single person)

    Recommends · Apr 2026 · in “Claude Opus 4.7 costs 20–30% more per session”

19. It's also a fraction of the cost of a Claude Code subscription especially when you account for how high the usage limits are.

    Recommends · Jun 2026 · in “Anthropic Accuses Alibaba of 'Illicitly' Accessing AI Models”

20. GLM 5.2, deepseek 4 Flash and the newly released Hy3 are Opus 4.8 and Sonnet 4.6 level models at a tiny fraction of the cost.

    Recommends · Jul 2026 · in “Anthropic's Method to Losing Goodwill in a Few Easy Steps”

21. Models in the 700B+ category (GLM5, Kimi K2.5) are decent, but running those on your own hardware is a six-figure investment.

    Mixed · Mar 2026 · in “Elevated Errors in Claude.ai”

22. For tasks that don't require vision I personally like the NVFP4 quant of GLM 5.3 from Local Inference Lab better than DS4.1F

    Recommends · Oct 2026 · in “Why isn't the industry freaking out about DeepSeek 4.1 Flash?”

23. I never hit quota limits. At work I get Claude, and I was getting reports that I was spending $75 per hour of work on Opus.

    Recommends · Jul 2026 · in “Open-weight AI is having its Kubernetes moment. Let's not ruin it”

24. Glm5.3 is my daily driver and I feel im having the most productive experience with agentic collaborations so far, by a lot.

    Recommends · Sep 2026 · in “Fable 5 – Median thinking declined in August”

25. So GLM 5.2 is more expensive then GLM 5.1 on their service (those thinking tokens), compared to API costs, its dirt cheap.

    Mixed · Jul 2026 · in “GLM 5.2 and the coming AI margin collapse”

26. 5.2 is a great model, and I use it for a lot of things, but it's noticeably below Opus 4.8 or GPT-5.5 in real-world usage.

    Mixed · Jul 2026 · in “Kimi K3 is now live”

27. I can confirm GLM 4.7 is possible on M1 Max and it can do nice comprehensive answers (albeit at 12 min an answer) locally.

    Mixed · Mar 2026 · in “Can I Run AI locally?”

28. anyone remotely considering that subscription should go on reddit and see all the people experiencing outages constantly and insanely slow speeds

    Criticizes · Apr 2026 · in “Reallocating $100/Month Claude Code Spend to Zed and OpenRouter”

29. A friend had GLM 5.2 suddenly become dumb for hours. It started forgetting instructions, randomly inserted some bits in simplified chinese

    Criticizes · Jul 2026 · in “Rethinking Legal Education in the AI Era”

30. GLM 5.3 knows that its implementation violates user requirements yet sticks with it after imagining what a hypothetical grader will check

    Criticizes · Sep 2026 · in “Speculative Reward Hacking in Coding Agents”

31. GLM 5.3 was more expensive than Haiku and Sonnet 5.5 on a workload I was testing recently despite being much cheaper on the price sheet

    Criticizes · Oct 2026 · in “Token-maxxing is dead, OpenRouter's token usage plateaus”

32. GLM and Kimi models are "thinkslop" models that reason over their own thinking, which increases cost and decreases speed significantly

    Criticizes · Aug 2026 · in “Qwen3.8-27B”

33. I have a hard time justifying GLM 5.3 these days. It's slightly better than Flash but rarely enough to justify the much steeper price.

    Criticizes · Oct 2026 · in “One month coding with GLM 5.3 Flash”

34. Typical case: 30% invalid output, every trial timed out, and two identical runs could produce a valid solution or complete garbage.

    Criticizes · Feb 2026 · in “GLM-5 topped the coding benchmarks. Then I used it”

35. I can show you my git tree where GLM4.5 deleted my whole test suite and my session docs where it invented new github cli commands.

    Criticizes · May 2026 · in “GLM-5V-Turbo: Toward a Native Foundation Model for Multimodal Agents”

36. a similarly bad result trying to run GLM 5.2 which fits in memory but the custom kernels we had to try to contrive were too slow

    Criticizes · Jun 2026 · in “Qwen 3.6 27B is the sweet spot for local development”

37. The biggest problem with Z.ai model for a long time is not quality, but the inference speed and general capacity availability.

    Criticizes · Jun 2026 · in “GLM 5.2 vs. Opus”

38. I dont even bother asking opus those kind of questions when I know deepseek-4.1-flash or glm-5.3* or mistral-large-4 will give the proper solution

    Recommends · Oct 2026 · in “I turned my unused Kindle into a photo frame with GPT-6 Astra”

39. Frankly, a lot of times I prefer using GLM 4.6 running on Cerebras Inference, than having to deal with the performance hiccups from Claude.

    Recommends · Jan 2026 · in “Apple picks Google's Gemini to power Siri”

40. That's mostly about 1) harness incompatibility with the GLM and/or 2) Bad implementation of hosting the model by your upstream vendor

    Recommends · Aug 2026 · in “GLM-5.3: Frontier Coding with Emergent Cyber Capabilities”