Loading video...

Video Failed to Load

Go Home

Today, we're introducing GLM-5.2 Fast: our GLM-5.2 Model API designed for the most demanding real-time use cases. The Fast tier delivers 2-3x higher TPS than the standard GLM-5.2 MAPI.

baseten's profile picture

Baseten

17,342 subscribers

114,533 views • 24 days ago •via X (Twitter)

Science & Technology
LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

GLM 5.2 Easy Setup with Baseten + Droid harness. 00:00 Fastest GLM 5.2 Setup Goal 00:32 Getting a Baseten API Key 01:09 Automated Droid AI Configuration 02:23 Testing the GLM 5.2 Integration 03:07 Alternative Setups and Limitations 04:14 Bonus: Setup for Other IDEs
5:15

GLM 5.2 Easy Setup with Baseten + Droid harness. 00:00 Fastest GLM 5.2 Setup Goal 00:32 Getting a Baseten API Key 01:09 Automated Droid AI Configuration 02:23 Testing the GLM 5.2 Integration 03:07 Alternative Setups and Limitations 04:14 Bonus: Setup for Other IDEs

Ray Fernando

10,580 views • 1 month ago

GLM-5.2 delivers a substantial leap in app development capabilities, which also represent demanding long-horizon tasks. Results: - GLM-5.1: 21/70 - GLM-5.2: 48/70 - Claude Fable 5: 56/70 That's more than a twofold improvement from GLM-5.1 to GLM-5.2. These come from an internal benchmark of 35 challenging mobile development tasks, each run twice for a total of 70 trials. We measured task completion, defined as core features working without major issues.
0:45

GLM-5.2 delivers a substantial leap in app development capabilities, which also represent demanding long-horizon tasks. Results: - GLM-5.1: 21/70 - GLM-5.2: 48/70 - Claude Fable 5: 56/70 That's more than a twofold improvement from GLM-5.1 to GLM-5.2. These come from an internal benchmark of 35 challenging mobile development tasks, each run twice for a total of 70 trials. We measured task completion, defined as core features working without major issues.

Zixuan Li

335,067 views • 1 month ago

we asked GLM 5.2 and Claude Opus 4.8 to find and fix the bugs in two different scripts while Opus 4.8 finished a few seconds faster in both tests, GLM 5.2 burned roughly 17k–38k fewer tokens and absolutely crushed it on cost, coming in 15x to 22x cheaper! Test 1: GLM 5.2 ($0.068) vs. Opus 4.8 ($1.06) Test 2: GLM 5.2 ($0.045) vs. Opus 4.8 ($1.02)
0:23

we asked GLM 5.2 and Claude Opus 4.8 to find and fix the bugs in two different scripts while Opus 4.8 finished a few seconds faster in both tests, GLM 5.2 burned roughly 17k–38k fewer tokens and absolutely crushed it on cost, coming in 15x to 22x cheaper! Test 1: GLM 5.2 ($0.068) vs. Opus 4.8 ($1.06) Test 2: GLM 5.2 ($0.045) vs. Opus 4.8 ($1.02)

GMI Cloud

86,444 views • 1 month ago

Grok 4.5 operates at Opus-level intelligence and Sonnet-level pricing! We tested it against Claude Opus 4.8, Sonnet 5, and GLM 5.2 on a SpaceX mission: liftoff, stage sep, orbit insertion, satellite deploy. Grok 4.5 animates it more accurately and vividly than Sonnet 5 and GLM 5.2. Opus is the fastest and most expensive. Grok burns more tokens but still lands ~40% cheaper than Opus, and a bit under Sonnet 5. Time / tokens / cost: Grok 4.5 — / 35k / $0.69 Opus 4.8 — / 13.7k / $1.10 Sonnet 5 — / 22.1k / $0.78 GLM 5.2 — / 92.7k / $0.13 yet GLM did the whole mission for $0.13 😮
0:45

Grok 4.5 operates at Opus-level intelligence and Sonnet-level pricing! We tested it against Claude Opus 4.8, Sonnet 5, and GLM 5.2 on a SpaceX mission: liftoff, stage sep, orbit insertion, satellite deploy. Grok 4.5 animates it more accurately and vividly than Sonnet 5 and GLM 5.2. Opus is the fastest and most expensive. Grok burns more tokens but still lands ~40% cheaper than Opus, and a bit under Sonnet 5. Time / tokens / cost: Grok 4.5 — / 35k / $0.69 Opus 4.8 — / 13.7k / $1.10 Sonnet 5 — / 22.1k / $0.78 GLM 5.2 — / 92.7k / $0.13 yet GLM did the whole mission for $0.13 😮

GMI Cloud

150,679 views • 1 month ago

I just tested GLM 5.2, Claude opus 4.8, and GPT 5.5 to see which one could rebuild my personal site in that whimsical ascii style GLM 5.2 held its own and went toe to toe with the other 2 top tier models lowkey impressed, look at the hero section it built here
1:36

I just tested GLM 5.2, Claude opus 4.8, and GPT 5.5 to see which one could rebuild my personal site in that whimsical ascii style GLM 5.2 held its own and went toe to toe with the other 2 top tier models lowkey impressed, look at the hero section it built here

Ann Nguyen

33,577 views • 2 months ago

We ran GLM 5.2 and Kimi K3 on 100 deep research tasks (DRACO by perplexity) and asked Fable to be the judge Kimi K3: 71.6 mean score, 77% pass rate GLM 5.2: 41.5 mean score, 22% pass rate K3 took 3h 38m and $60. GLM finished in 1h 45m for $12.50. Deep Dive in 🧵
0:30

We ran GLM 5.2 and Kimi K3 on 100 deep research tasks (DRACO by perplexity) and asked Fable to be the judge Kimi K3: 71.6 mean score, 77% pass rate GLM 5.2: 41.5 mean score, 22% pass rate K3 took 3h 38m and $60. GLM finished in 1h 45m for $12.50. Deep Dive in 🧵

GMI Cloud

80,888 views • 24 days ago

I love using GLM 5.2 for web app iteration. My workflow: generate 6 variations, then pick the best one and continue iterating on it. I built Recast to make this even easier. Give it a prompt, get 6 variations, download the code for your favorite version, then continue iterating in your own coding agent. GLM 5.2 works really well for this use case since it's fast, cheap, and great at design. On average, you can make 3-6 landing pages with GLM 5.2 for every one you can generate with Opus 4.8. And the design variety of the GLM landing pages helps a lot with trying to think through what style you want to go with. Recast is launching soon (as a web app + skill).
0:34

I love using GLM 5.2 for web app iteration. My workflow: generate 6 variations, then pick the best one and continue iterating on it. I built Recast to make this even easier. Give it a prompt, get 6 variations, download the code for your favorite version, then continue iterating in your own coding agent. GLM 5.2 works really well for this use case since it's fast, cheap, and great at design. On average, you can make 3-6 landing pages with GLM 5.2 for every one you can generate with Opus 4.8. And the design variety of the GLM landing pages helps a lot with trying to think through what style you want to go with. Recast is launching soon (as a web app + skill).

Hassan

20,164 views • 1 month ago

Opus 4.8 vs GLM 5.2 everyone is praising Glm on my timeline, so i had to test it myself > both models were given the same prompt > GLM was super fast and used like 2% of my weekly limit > Opus took over 20 mins and used like 3% of weekly limits which one do you think won?
0:41

Opus 4.8 vs GLM 5.2 everyone is praising Glm on my timeline, so i had to test it myself > both models were given the same prompt > GLM was super fast and used like 2% of my weekly limit > Opus took over 20 mins and used like 3% of weekly limits which one do you think won?

J A Z I I

48,108 views • 1 month ago

GLM-5.2 running under my desk, modifying GGUFs to enable MTP on the Spark. What a world we live in, where you can have a megamind running off electrons going really fast.
0:27

GLM-5.2 running under my desk, modifying GGUFs to enable MTP on the Spark. What a world we live in, where you can have a megamind running off electrons going really fast.

0xSero

23,422 views • 1 month ago

Timelapse #156 (36 hrs) - Worked with the tiny corp on getting GLM 5.2 running on 8xMI300X (sglang won here) - Launched KernelBench-Mega and updated Kernelbench-Hard with h100 and b200 sweeps - Took care of boring business stuff - Did some training sweeps for specialized technical vocab audio model - Some bugs with putting kernelbench-mega and hard on cloud instances so had to do some reruns. Learned a lot though - Setting up my own local rl infra and profiling concurrency 128 rollouts with vllm. Became clear to me that I need to serve in nvfp4, use MoE only for throughput, reap so training doesn’t OOM, dig into the vllm kernel graph itself to not underutilize my hardware from poor flashinfer/cutlass selections for my rtx pro 6000 sm120 architecture - Might do online distillation from glm 5.2 but for now taking it one step at a time - Slept for a bit then woke up and showered - Fixed an issue with SGLang tensor parallel deadlock on GLM 5.2 architecture with MTP enabled - GLM 5.2 inference is 2-3x faster than coding plans and running on amd boxes - Spent time with family - Hung out with some friends - Recorded some yoctogpt lectures with the revamped notebook (high taste btw) - Setting up dflash training for GLM 5.2
0:30

Timelapse #156 (36 hrs) - Worked with the tiny corp on getting GLM 5.2 running on 8xMI300X (sglang won here) - Launched KernelBench-Mega and updated Kernelbench-Hard with h100 and b200 sweeps - Took care of boring business stuff - Did some training sweeps for specialized technical vocab audio model - Some bugs with putting kernelbench-mega and hard on cloud instances so had to do some reruns. Learned a lot though - Setting up my own local rl infra and profiling concurrency 128 rollouts with vllm. Became clear to me that I need to serve in nvfp4, use MoE only for throughput, reap so training doesn’t OOM, dig into the vllm kernel graph itself to not underutilize my hardware from poor flashinfer/cutlass selections for my rtx pro 6000 sm120 architecture - Might do online distillation from glm 5.2 but for now taking it one step at a time - Slept for a bit then woke up and showered - Fixed an issue with SGLang tensor parallel deadlock on GLM 5.2 architecture with MTP enabled - GLM 5.2 inference is 2-3x faster than coding plans and running on amd boxes - Spent time with family - Hung out with some friends - Recorded some yoctogpt lectures with the revamped notebook (high taste btw) - Setting up dflash training for GLM 5.2

Elliot Arledge

45,447 views • 1 month ago

GLM 5.2 just beat Fable 5 at website design. 🌐 The crazy part: GLM is text-only. It can build the site, but it can’t inspect the result. So we paired it with Browser Use v2 multimodal QA subagents. > Reviews the site and finds bugs > Judges the aesthetics > Sends GLM targeted fixes to iterate on Build + full QA cost less than $0.75. Try it now ↓
0:25

GLM 5.2 just beat Fable 5 at website design. 🌐 The crazy part: GLM is text-only. It can build the site, but it can’t inspect the result. So we paired it with Browser Use v2 multimodal QA subagents. > Reviews the site and finds bugs > Judges the aesthetics > Sends GLM targeted fixes to iterate on Build + full QA cost less than $0.75. Try it now ↓

Browser Use

273,561 views • 1 month ago

JUST IN: 🇨🇳 China’s new open-source AI model, GLM-5.2 by Zhipu AI, reportedly matches Anthropic’s Claude Mythos in security bug detection. The model outperformed some Claude versions in independent cybersecurity benchmarks and can achieve Mythos-level performance. Unlike Claude Mythos, GLM-5.2 is open-source and significantly cheaper to run. The gap between US and Chinese AI models is narrowing rapidly.
0:32

JUST IN: 🇨🇳 China’s new open-source AI model, GLM-5.2 by Zhipu AI, reportedly matches Anthropic’s Claude Mythos in security bug detection. The model outperformed some Claude versions in independent cybersecurity benchmarks and can achieve Mythos-level performance. Unlike Claude Mythos, GLM-5.2 is open-source and significantly cheaper to run. The gap between US and Chinese AI models is narrowing rapidly.

CryptoTweets

62,870 views • 1 month ago

Introducing GLM 5.2 for autoresearch GLM 5.2 is the first open weights model we've tried on our autoresearch pipeline that's proven capable for real research tasks. With Fable 5's restrictions on research, having an open weights alternative is a huge win for open source Watch it carry out fully async vs colocated sync RL training on Harbor code contests across two 8xH100 nodes on top of SkyRL. Resolves setup issues, tracks runs to completion, and produces a full comparison of throughput and reward stability
0:48

Introducing GLM 5.2 for autoresearch GLM 5.2 is the first open weights model we've tried on our autoresearch pipeline that's proven capable for real research tasks. With Fable 5's restrictions on research, having an open weights alternative is a huge win for open source Watch it carry out fully async vs colocated sync RL training on Harbor code contests across two 8xH100 nodes on top of SkyRL. Resolves setup issues, tracks runs to completion, and produces a full comparison of throughput and reward stability

alphaXiv

271,138 views • 1 month ago

CHINA IS CATCHING UP IN AI 🇨🇳 China’s new open-source AI model, GLM-5.2 by Zhipu AI, reportedly matches Anthropic’s Claude Mythos in security bug detection. The model outperformed some Claude versions in independent cybersecurity benchmarks and can achieve Mythos-level performance. Unlike Claude Mythos, GLM-5.2 is open-source and significantly cheaper to run. The gap between US and Chinese AI models is narrowing rapidly.
0:29

CHINA IS CATCHING UP IN AI 🇨🇳 China’s new open-source AI model, GLM-5.2 by Zhipu AI, reportedly matches Anthropic’s Claude Mythos in security bug detection. The model outperformed some Claude versions in independent cybersecurity benchmarks and can achieve Mythos-level performance. Unlike Claude Mythos, GLM-5.2 is open-source and significantly cheaper to run. The gap between US and Chinese AI models is narrowing rapidly.

CryptoGoos

71,944 views • 1 month ago

This is GLM 5.2 creating and describing *images* How, since the model is text-only? With Amp's specialized subagents. Multi-modal agentic workflows at the pareto frontier of intelligence, speed, and cost.
0:23

This is GLM 5.2 creating and describing *images* How, since the model is text-only? With Amp's specialized subagents. Multi-modal agentic workflows at the pareto frontier of intelligence, speed, and cost.

Beyang

20,111 views • 1 month ago

GLM-5.2 is legit. It produces games easily and as accurately as SOTA closed-source models. My GLM 5.2 versus Fable 5 comparison is close, but I prefer Fable 5 just a little. Both games are equal in nearly all respects, but Fable 5 produced slightly better graphics and more challenging gameplay. Both needed 2 shots to get all the prompt instructions correct.
1:00

GLM-5.2 is legit. It produces games easily and as accurately as SOTA closed-source models. My GLM 5.2 versus Fable 5 comparison is close, but I prefer Fable 5 just a little. Both games are equal in nearly all respects, but Fable 5 produced slightly better graphics and more challenging gameplay. Both needed 2 shots to get all the prompt instructions correct.

David Hendrickson

18,779 views • 1 month ago

Ran two prompts in Kimi K2.7 Code HighSpeed, Kimi K2.7 Code, and GLM 5.2 HighSpeed is 2.5 to 4.2 times faster than Kimi K2.7 Code while producing comparable quality 4.5 to 15 times faster than GLM 5.2, and half the price (or less) These are simple tests, so the pricing difference is small and speed difference could diverge a lot But definitely a hell of a lot faster with good output
0:53

Ran two prompts in Kimi K2.7 Code HighSpeed, Kimi K2.7 Code, and GLM 5.2 HighSpeed is 2.5 to 4.2 times faster than Kimi K2.7 Code while producing comparable quality 4.5 to 15 times faster than GLM 5.2, and half the price (or less) These are simple tests, so the pricing difference is small and speed difference could diverge a lot But definitely a hell of a lot faster with good output

GMI Cloud

38,174 views • 1 month ago

Holy Smokes GLM 5.2 beats Opus in many webdev task😳 This is the best open model , according to jietang said FABLE like model from them before Q1 2027 . i can trust them , just make glm fat with vision we will have fable Left Right
0:13

Holy Smokes GLM 5.2 beats Opus in many webdev task😳 This is the best open model , according to jietang said FABLE like model from them before Q1 2027 . i can trust them , just make glm fat with vision we will have fable Left Right

Chetaslua

22,917 views • 2 months ago

Loading DeepSeek v4 Flash just reminded me how good and fast this model is and how well it runs on DGX Sparks. GLM-5.2 is really awesome, but if you want fast and concurrent inference, DS4F is still a top choice IMO. Created by DS4F in single shot The coffee guy
0:20

Loading DeepSeek v4 Flash just reminded me how good and fast this model is and how well it runs on DGX Sparks. GLM-5.2 is really awesome, but if you want fast and concurrent inference, DS4F is still a top choice IMO. Created by DS4F in single shot The coffee guy

Mia

39,686 views • 20 days ago

Chinese labs aren't just dropping model weights anymore, they're building full ecosystems. Z.ai's ZCode + GLM 5.2 = native IDE, 1M context, agentic workflows, and now dSpark speculative decoding. The open model platform war is here. 👇
7:19

Chinese labs aren't just dropping model weights anymore, they're building full ecosystems. Z.ai's ZCode + GLM 5.2 = native IDE, 1M context, agentic workflows, and now dSpark speculative decoding. The open model platform war is here. 👇

Boxmining

14,831 views • 1 month ago