I did some model hacks, and got GLM5.2 from about 2.5 tok/s to >50 tok/s on my GH200 system.
G'day. This is part 3 on my Local LLM adventures. I have a crazy system hacked server-to-desktop system: Component Spec GPUs 2x Hopper H100, 96 GB…
btw Zai IPO'ed in Jan at HK$120 a share. when I first met @louszbd nobody really knew anyone using GLM's. now they have beat deepseek with the world's…
btw Zai IPO'ed in Jan at HK$120 a share. when I first met @louszbd nobody really knew anyone using GLM's.now they have beat deepseek with the…
RT Hassan: Ran 10 more tests comparing GLM 5.2 & Opus. On average, GLM 5.2 produced 2x the tokens but was still faster + 3x cheaper with similar quali…
RT HassanRan 10 more tests comparing GLM 5.2 & Opus.On average, GLM 5.2 produced 2x the tokens but was still faster + 3x cheaper with similar…
Is it possible to run a giant model like GLM5.2 on this cluster (4x servers with 512GB RAM + dual AMD Epyc)? 16 channel memory should hit 409GB/s per node.
Hey all, I have a piece of hardware laying around which is pretty fast from a traditional (non-GPU) server viewpoint. The hardware is the following: Dell…
GLM-5.2 OpenAI-Compatible API: A Hands-On Guide to Reasoning Effort, Function Calling, and Long-Context Retrieval
We build a practical GLM-5.2 workflow using its hosted, OpenAI-compatible API instead of running the model locally. We set up multiple providers, load the API key…
RT Zixuan Li: GLM-5.2 is available in Perplexity's Agent API. Just tested it, and it's powerful when paired with the Search SDK inside a sandbox. – Sp…
RT Zixuan LiGLM-5.2 is available in Perplexity's Agent API. Just tested it, and it's powerful when paired with the Search SDK inside a sandbox.- Spin up…
RT afra wang: thing i know about http://z.ai, the company behind GLM: 1. http://z.ai is known as "Zhipu" before being rebranded as a sleeker "http://z…
RT afra wangthing i know about http://z.ai, the company behind GLM:1. http://z.ai is known as "Zhipu" before being rebranded as a sleeker "http://z.ai". The Chinese name…
RT Hassan: Introducing The Blind Test. Two landing pages. One built by GLM 5.2 and one by Opus 4.8. Can you tell which is which? It's very difficult t…
RT HassanIntroducing The Blind Test.Two landing pages. One built by GLM 5.2 and one by Opus 4.8.Can you tell which is which?It's very difficult to get…
GLM-5.2 should be “DeepSeek moment” for agents. We enter a new world where the top end of agentic capabilities are available in open models. If you …
GLM-5.2 should be “DeepSeek moment” for agents. We enter a new world where the top end of agentic capabilities are available in open models.If you care…
GLM-5.2 is the step change for open agents
A capability threshold I've been carefully monitoring.
RT Carol Lin: GLM 5.2 x AWS GLM-5.2 is accessible via http://Z.ai GLM API on AWS Marketplace 🚀 Powerful long-horizon autonomous workflows, top-tier…
RT Carol LinGLM 5.2 x AWSGLM-5.2 is accessible via http://Z.ai GLM API on AWS Marketplace 🚀Powerful long-horizon autonomous workflows, top-tier coding & multi-step agent reasoning capability,…
An hour in and first impression is definitely that GLM is really solid (very easy to set up on @FireworksAI_HQ, props to them for that, took me like 5…
An hour in and first impression is definitely that GLM is really solid (very easy to set up on @FireworksAI_HQ, props to them for that, took…
Open weights models, via GLM 5.2, had their "very practically useful" in coding harness moment before Gemini. ~200 days since the release of Opus 4.5.
Open weights models, via GLM 5.2, had their "very practically useful" in coding harness moment before Gemini. ~200 days since the release of Opus 4.5.
A year ago this would have been an obvious closed-model task. Now GLM-5.2 can read the issue, reason through the scene, patch the code, and keep movin…
A year ago this would have been an obvious closed-model task.Now GLM-5.2 can read the issue, reason through the scene, patch the code, and keep moving…
GLM-5.2 on Together AI is showing up fast on @OpenRouter ⚡️ The model is strong, and our serving path makes that strength usable in the loop. Togeth…
GLM-5.2 on Together AI is showing up fast on @OpenRouter ⚡️The model is strong, and our serving path makes that strength usable in the loop.Together has…
RT ⚡AI Search⚡: GLM 5.2 continues to impress me. Here's its result on Vending Bench, which measures an AI's performance on running a business over a…
RT ⚡AI Search⚡GLM 5.2 continues to impress me. Here's its result on Vending Bench, which measures an AI's performance on running a business over a long…
RT Itamar Golan 🤓: Hot take: GLM 5.2 might be the first open/public model that actually changes the enterprise AI cost equation. I played with it f…
RT Itamar Golan 🤓Hot take:GLM 5.2 might be the first open/public model that actually changes the enterprise AI cost equation.I played with it for a few…
RT Thomas Wolf: Desert island survival list: ✅ Solar panel / battery ✅ 256 GB Mac Studio ✅ GLM 5.2 Civilization in a backpack
RT Thomas WolfDesert island survival list:✅ Solar panel / battery✅ 256 GB Mac Studio✅ GLM 5.2Civilization in a backpack
RT Lasse: glm 5.2 is awesome
RT Lasseglm 5.2 is awesome
RT Dan Loewenherz: GLM-5.2 is really, really good.
RT Dan LoewenherzGLM-5.2 is really, really good.
RT Alex Brandes ²: GLM 5.2 is really good. Wow.
RT Alex Brandes ²GLM 5.2 is really good. Wow.
RT Niels Rogge: Here's how to use Claude Code with GLM-5.2 via @huggingface Inference Providers: 1. Create a token at https://huggingface.co/settings/…
RT Niels RoggeHere's how to use Claude Code with GLM-5.2 via@huggingface Inference Providers: 1. Create a token at https://huggingface.co/settings/tokens (fine-grained, enable "Make calls to Inference Providers")…
[AINews] GLM > GPT? GLM-5.2 passes vibe check; Z.ai forecasts Open Fable by December
With GLM-5.2 passing everyone's vibe check, the open models story finally becomes a real frontier story.
Really looking forward to one of the super-fast custom silicon inference providers like @GroqInc or @cerebras getting GLM 5.2 running Cerebras has GLM…
Really looking forward to one of the super-fast custom silicon inference providers like @GroqInc or @cerebras getting GLM 5.2 runningCerebras has GLM-4.7, Groq is still mostly…