GLM-5.2: Open Weights, Near-Frontier Intelligence — Zixuan Li, Z.ai

Video authored by the channel shown, not by IANIA; all rights belong to the creator. Transcript and summary generated automatically; the original-language content is preserved.

Original language: English

Original video: GLM-5.2: Open Weights, Near-Frontier Intelligence — Zixuan Li, Z.ai · Channel: AI Engineer · 13:12 · Watch on YouTube

Topics: AI agents, AI models, Local and open-source models, Open source and software ecosystems

Summary

Z.ai’s keynote introduces GLM 5.2 and explains how the GLM name evolved from an early language-model research term into the company’s model brand. The speaker presents GLM 5.2 as a major open-weight model with stronger reasoning, coding, agentic, and general-purpose capabilities, including improved performance without a thinking budget. He describes open weights as a way to meet needs for security, on-premise use, fine-tuning, and collaboration with the wider ecosystem. The talk also points to resources for trying GLM 5.2 and announces Z code, a coding harness that supports GLM and other frontier models.

Key points

Zhipu is the company, while GLM is the name used for its models; the name originated with the 2021 General Language Model pre-training paper.
The speaker frames intelligence as broader than IQ-style tests, including reasoning, coding, and agentic capabilities.
GLM 5.2 is presented as a substantial improvement over earlier versions on difficult and long-horizon tasks.
The speaker says the non-thinking model outperforms the GLM 5.1 thinking model, while GLM 5.2 also offers a high thinking-budget level.
GLM is intended for general chat, math, role play, and daily workflows, not only coding.
Open weights can support security, control, trust, and on-premise deployment, including for enterprise and government users.
Open weights also let organizations fine-tune models for specialized areas such as legal, finance, and security.
The speaker credits open-source software contributors, model users, and application builders with helping improve GLM.
The GLM 5.2 tech blog includes links to the model, ways to try it, API access, a coding plan, and information about its training pipeline.
Z code is introduced as a coding harness built for GLM 5.2 that also supports other frontier models.

Chapters

Start	Chapter	Summary
00:12	Introduction and GLM background	The speaker introduces Z.ai and GLM 5.2, clarifies the relationship between Zhipu and the GLM name, and outlines the company’s model history.
03:23	GLM 5.2 capabilities	The keynote describes progress in reasoning, coding, and agentic tasks, as well as improvements in thinking budgets and general capabilities.
06:04	Why release open weights	The speaker explains the value of GLM’s broader use beyond coding and gives reasons for opening model weights, including deployment, fine-tuning, and collaboration.
09:22	Community resources and Z code	The talk thanks open-source contributors, points viewers to GLM 5.2 resources, and announces the Z code harness.
12:07	Closing remarks	The host thanks the speaker and describes the event’s role in bringing AI companies, labs, and ecosystem partners together.

Quotes

GLM actually represent general language model pre-training with auto-regressive blank filling. — 02:17

the non-thinking model is better than the 5.1 thinking model. — 05:29

GLM is more than coding model — 06:40

we open the weights for several things because there are users’ needs and there are our own needs. — 07:13

we actually have our own harness. — 11:00

Concepts and entities

GLM-5.2, GLM-5.1, GLM, Zhipu, Z.ai, Zixuan Li, Z Code, Hugging Face, open-weight models, fine-tuning

Full transcript (original language)

00:12 Without further ado, we’re going to bring on Zishan for for his keynote if we’re all set. Hey, how are you? >> Great I’m here. >> Yes, to the magic of the internet. It’s really good to see you again. Zishan has you’ve been speaking with AI Singapore and obviously GLM is the talk of the town right now. There’s been an amazing booth downstairs in the expo and I think a lot of people are just very excited to hear from you on what all is going on with ZIAI. So if

00:44 you want to take it away. >> Yes, sorry again for not showing up in person, but I have a team whole team coming to the town and we have a booth. So you have any questions, feel free to reach out to me on X link and also you can look for my team on our booth. And since I cannot see my slides, so I will need Swee to help me flip through all the slides for me. Yeah. >> You’re good. >> Here Swee, like what which slides we are

01:16 in right now? >> Yeah, so we’re we’re on the opening slide. We’re talking about intelligence and ZIAI. >> Okay, so yeah, so you can you can see that it’s the first time for us to introduce GLM 4.2 5.2 to the world and also we are going to share something about Z.AI and GLM because maybe people will think Z.AI and GLM they’re irrelevant, right? You are your company is not G.AI or your

01:47 model’s called Z1 or Z2. And you can find my X and our company’s X account here, so you can just search for my name Zishan Lee and the Z A I org. And you can follow them for the follow-ups. So, yeah, we can go to the second slide. Yeah, actually I cannot see the slide, so I’ll try to >> So, we we >> Yeah, try to make sure that it is correct. So, the company actually is

02:17 called Zhipu. Maybe some people have heard of it. And the all the models, it’s it’s called GLM. Actually, it’s not a a brand name, it’s a generic term. So, GLM actually represent general language model pre-training with auto-regressive blank filling. And that paper was published back in 2021. So, actually we were the one of the first labs to do explorations on large

02:47 language models at the same time with OpenAI, and Anthropic, and DeepMind. And even today, we we no longer use GLM as the architecture, we still use the name GLM as our brand name. So, we use GLM 5.1, 5.2, and it become like one of our like most prod uh proddest product and model. And the second thing that we look for is intelligence upper bound. So, in terms of intelligence, we we may feel that it’s represent IQ, something something

03:23 like that. And when Deep Seek launched one 01 launch, people are talking about the model’s capability to solve math problems, physics problems. But what actually intelligence mean is not just like IQ or AIME or other other physics problem. So, from GLM 4.5 to GLM 4.7, we are exploring like several things, like reasoning, coding, and agentic capabilities. So, as you can see from the slides, So, we add like

03:53 Yeah, the last slide. Yeah, we were actually I haven’t like finished that slide. Yeah, so Okay. Yeah, need to go back to the two two slides. >> I don’t have back I don’t have a back button. Thank you. I would you I don’t understand clickers that don’t have back buttons. Like why Okay, I you know, anyway, go ahead.

04:24 >> Yeah, never mind. Yeah, because people want to see the GLM 5.2. They they don’t want to see you like GLM 4.1 or 5. But like GLM 4.2 actually specialize in coding and genetic task as you can see from the graph because there are a lot of rumors whether your model is close to mythos, fable, but actually I want to share these slides to all of you. So, you can see it’s somewhere between Opus 4.7 and 4.8 and we use the hardest problems like Deep Sweep, Terminal Bench 211, which was mentioned by the Open AI

04:57 team uh several minutes ago. And all the like long horizon task and benchmark shows that the capabilities are on par with at least Opus 4.7 and it it it shows a significant improvements over 1.1. Also, for GLM 1.2, we add a thinking uh level called high. So, because we also noticed as we move to the harder task it may

05:29 consume more tokens and also we we care a lot about the token efficiency. So, it’s the first time we add the high level for thinking budget. But even without thinking, the non-thinking model is better than the 5.1 thinking model. So, I think it’s a huge improvement for the open weight model. That’s what really impressed the world and why people are talking about GLM 5.2 lately. Okay, the next slide. And one one thing that I want to mention is that GLM is more more than coding

06:04 model because people use it inside clock code code X, open code. But actually, we have trained a lot of things outside coding. For example, we improve a lot in GDP valve and also math problems. We also care about math problems, frankly speaking. And also, we train a lot of thing uh related to role play, general chat. We want to improve every aspect of the model. So, you can see from the artificial analysis intelligence index actually leads the other open way model a lot and close to

06:40 the frontier model. So, I want you if you are you haven’t experienced GLM yet, you can use GLM to do general chat, use it to process your daily workflow, not just for coding, but you can explore the model like beyond the coding scope. Next slide. And GLM 4.2 is a open weight model. So, people always ask me, “Why do you open weight? So, do you care about your business or like do you care about

07:13 losing market to some infra suppliers?” But actually, we open the weights for several things because there are users’ needs and there are our own needs. If we can meet their needs, I think it’s okay. It’s definitely okay for us to to open the model. For example, if our users want security and control and we want to build trust, we can open weight model. For for some cases, if enterprise or government

07:44 especially in the Western world, want to use the model, we open way, we upload to the hugging face so that they can use the model on premise. I think it’s very beneficial for the whole ecosystem it to explore the model, especially when the capabilities is close to the frontier model. And second, if people want diversity. So, in terms of diversity, I mean the capabilities in legal, finance, security, they can fine-tune the model.

08:15 So, we need to open the model to let them fine-tune. For example, Harvey is fine-tuning GLM 4.1, and maybe they’re thinking about fine-tuning GLM 5.2 afterwards. And I’ve I have like talked to a lot of other companies. They’re also thinking about fine-tuning GLM as their like next step or next strategy to differentiate themselves from other application company. And third, if our customer or an individual want to co-design and

08:52 predict the future, sometimes they need to see the architecture of the model. They need to see the recipe of how you train the model. So, we want to make the norm. We want to make my bets. So, we want to co-shape the future with our customers with the open source uh community. So, I think our needs and our ecosystem really like fits into each other. And GLM 5.2 couldn’t succeed without

09:22 you, all the open source community players like Onflos and Media and and some like individuals, super developers, they are part of it. And thanks a lot to application builders like Peter. Right? Because open source doesn’t just include open source model, but also open source softwares and open source other sorts of support. So, all your support and what are you you’re doing

09:55 right now pushes to to make better models. I think uh you you are the true hero. And the last, I want to share a a great resource for you to go through. Geo 5.2 is our tech blog. So, actually in that tech blog, we share several things like our hugging face repo, how how you can try the Geo and 5.2. You can try the inside the chatbot agent, and also you can call the API. And also

10:27 we have a coding plan like the Codex or cloud core subscription for you to use your tokens as individuals. And also we share something about our training pipeline training recipe, which you can understand why it’s a good model. So, there are a lot of things behind the behind the model, not just um a a model that had great data. We also have fantastic technologies behind the model, so you can export the model yourself. You can see what difficulties

11:00 we have gone through. And the last slide, actually it’s kind of a one more thing. So, it’s the first time we share the code to the whole community. Woo! So, that one more thing is we actually have our own harness. The Z code, actually it’s built for Geo 5.2, but also support all frontier models. You can bring your own key. You can connect to Z code. Actually, the I think the operation is is similar to

11:33 Kodak. Actually, you you can try some techniques like go or other like compact technique techniques like what you’ve done in in Kodak and car code. And this harness is I think it’s the perfect one for GLM 102. If you haven’t experienced it, you can just search Z code or you can go to our booth. We have our team members showing this harness to you. And welcome to to our booth. And welcome to like talk to me in the future. And

12:07 next time I’ll definitely be in SF talking to everyone. Yeah, thanks. >> Thank you very much. Um So uh trust me, the very first World’s Fair, I wasn’t allowed to come back in the country for my own conference, so I know exactly this feeling. Uh but it it it’s all good. Uh the ZAI team, I really appreciate them uh making the effort. They really want to meet you. They’re here to meet you. And this is the whole point of the World’s Fair, to bring all the top world’s AI uh

12:41 companies and labs uh all in one place, so you can do business together, uh meet the people behind the models that you use, ask the questions that he cannot answer in public, but you can ask in private. Um I’m also very proud. Uh he showed uh Zhicheng showed that that list of uh Hugging Face uh you know, top uh contributors. I think we are four for eight uh in that list uh present at World’s Fair. Uh we’re working with Nvidia and Unsloth and Ollama and then uh all those other fine-tuners as well to uh to make our our sort of the

13:12 inference and the local tracks uh that you’re going to see over the next few days. Um with that, thank you so much to Zhicheng. Uh I’m going to invite on uh the next uh couple folks. I think I I I think I might be doing Ali’s job here, so uh we’ll we’ll talk to Hugging Face next and Minimax. Thank you.