/Tech2h ago

Z AI's 744-billion-parameter GLM-5.2 scores 51 on the Artificial Analysis Intelligence Index, ranking fifth overall

Story Overview

Z AI just dropped GLM-5.2, a 744-billion-parameter open-weights model that lands at 51 on the Artificial Analysis Intelligence Index. That score puts it fifth overall while beating Gemini 3.5 Flash and DeepSeek V4 Pro, showing open models can still punch above their weight on broad intelligence metrics.

90900100125101.8K

#573

Original post

Zephyr@zephyr_z9#1695inTech

LFG!!! 51 at AA @teortaxesTex

Artificial Analysis@ArtificialAnlys

Z ai’s GLM-5.2 is the new leading open weights model on the Artificial Analysis Intelligence Index scoring 51 and it sits on the Pareto frontier of Intelligence vs Cost per Task

@Zai_org’s GLM-5.2 is the same size as GLM-5.1 (744B total / 40B active parameters) but scores 11 points higher on the Intelligence Index v4.1, placing ahead of MiniMax-M3 (44) and DeepSeek V4 Pro (max, 44). On the first-party API it is priced in line with GLM-5.1 at $1.4/$4.4/$0.26 per 1M input/output/cache hit tokens

Key results:

➤ GLM-5.2 is the leading open weights model on the Intelligence Index v4.1. At 51, it leads MiniMax-M3 (44), DeepSeek V4 Pro (max, 44) and Kimi K2.6 (43)

➤ Improvements across most evaluations, particularly scientific reasoning: GLM-5.2 gains over GLM-5.1 on most evaluations, led by scientific reasoning on CritPt (+16 points to 21%) and HLE (+12 points to 40%), alongside AA-LCR (+9 points to 71%), tau3 banking (+15 points to 27%) and SciCode (+7 points to 50%). TerminalBench v2.1 also improves (+16 points to 78%) and GPQA Diamond gains 3 points to 89%

➤ Leading open weights model on GDPval-AA v2 and competitive with proprietary models: GLM-5.2 scores 1524 on GDPval-AA v2, ahead of MiniMax-M3 (1418) and DeepSeek V4 Pro (max, 1328). This impressive result places GLM-5.2 in-line with proprietary models including GPT-5.5 (xhigh reasoning). GDPval-AA v2 builds on the original GDPval-AA by baselining Elo to human performance at 1000, introducing a rotating panel of frontier-model judges, and raising the turn limit from 100 to 250 for longer-horizon agent trajectories

➤ GLM-5.2 uses more output tokens per task than other leading open weights models: the model uses 43k output tokens per Intelligence Index task, up from GLM-5.1 (26k) and above MiniMax-M3 (24k), Kimi K2.6 (35k) and DeepSeek V4 Pro (max, 37k)

➤ On the Intelligence vs. Cost per Task Pareto Frontier: GLM-5.2 is on the Pareto frontier of the Intelligence vs Cost per Task chart, with the lowest cost per task among models at its intelligence level. GLM-5.2 costs ~$0.46 per task, compared to GLM-5.1 ($0.25), Kimi K2.6 ($0.31), MiniMax-M3 ($0.18) and DeepSeek V4 Pro (max, $0.05)

Additional Model Details:

➤ License: MIT

➤ Size: 744B total parameters, 40B active parameters, equivalent to GLM-5.1

➤ Context window: 1M tokens, up from 200K on GLM-5.1

➤ Pricing: $1.4/$0.26/$4.4 per 1M input/cache hit/output tokens

➤ Availability: Alongside Z ai's first-party API, GLM-5.2 is available across third-party providers including @DeepInfra, @novita_labs, @nebiusai, @parasailnetwork , @SiliconFlowAI , @gmi_cloud , @Baseten and @FireworksAI_HQ

12:02 AM · Jun 17, 2026 · 16.9K Views

Benchmark Focus

Where the gains actually show up

The model pulls ahead on scientific reasoning and agent benchmarks compared with its own predecessor, though full details on every eval remain limited right after launch.

Open Weights

Who can run it today

Weights are already out under MIT license on Hugging Face, and the API is live for existing subscribers with the same per-token rates as the prior version.

Sentiment

Many users are excited about GLM-5.2's strong benchmark gains in intelligence and reasoning for open-weight models, while a few question the scoring system's reliability and efficiency tradeoffs.

Pos

91.2%

Neg

8.8%

19 comments with sentiment.

Cluster Engagement

Digg Deeper

No Digg Deeper questions have been answered for this story yet.

Posts from X

Most Activity

VIEWS3.3K

Artificial Analysis@ArtificialAnlys

Breakdown of the individual evaluations in the Artificial Analysis Intelligence Index v4.1

3h3.3K262

BOOKMARKS4LIKES57

Artificial Analysis@ArtificialAnlys

GLM-5.2 leads all open weights models on GDPval-AA v2, our primary metric for real-world agentic performance. At 1524 it places ahead of MiniMax-M3 (1418) and DeepSeek V4 Pro (max, 1328), and is effectively level with GPT-5.5 (xhigh, 1514). We visually inspected GLM-5.2's outputs across a range of GDPval-AA tasks. We have attached a selection below

3h3.3K574

RETWEETS93

Artificial Analysis@ArtificialAnlys

Z ai’s GLM-5.2 is the new leading open weights model on the Artificial Analysis Intelligence Index scoring 51 and it sits on the Pareto frontier of Intelligence vs Cost per Task

Key results:

➤ GLM-5.2 is the leading open weights model on the Intelligence Index v4.1. At 51, it leads MiniMax-M3 (44), DeepSeek V4 Pro (max, 44) and Kimi K2.6 (43)

Additional Model Details:

➤ License: MIT

➤ Size: 744B total parameters, 40B active parameters, equivalent to GLM-5.1

➤ Context window: 1M tokens, up from 200K on GLM-5.1

➤ Pricing: $1.4/$0.26/$4.4 per 1M input/cache hit/output tokens

3h91.5K794121

REPLIES3

Jonathan T.@JohnDotOwl

@ArtificialAnlys @Zai_org I thought the benchmark is reliable until i saw Gemini Flash , you got to be joking me right?

3h8346

Artificial Analysis@ArtificialAnlys

GLM-5.2 scores 4 on the AA-Omniscience Index, up from GLM-5.1 (2). The gain comes from both higher accuracy (25.1% vs 24.2%) and a lower hallucination rate (28.1% vs 29.4%), with attempt rate flat at 47%

3h2.5K29

Artificial Analysis@ArtificialAnlys

GLM-5.2 uses 43k output tokens per Intelligence Index task, of which 37k is reasoning. This is up from GLM-5.1 (26k) and higher than open weights peers MiniMax-M3 (24k) and Kimi K2.6 (35k), placing it among the less token-efficient open weights models at its intelligence level. GLM-5.2 sits off the most attractive quadrant on the Intelligence vs Output Tokens chart

3h2K231

Artificial Analysis@ArtificialAnlys

Compare GLM-5.2 with other leading models at: https://artificialanalysis.ai/models/glm-5-2

3h2.9K18

Zayd Honey@TheHoneyX

@ArtificialAnlys @Zai_org GLM is cooking I genuinely didn't expect it. Google really needs to step up. I am so tired of Gemini 3.1 pro, and 0 word responses or like empty responses it's so annoying

2h74041

Yiğit Polat@dyigitpolat

@ArtificialAnlys @Zai_org we also want to see where the Kimi K2.7-Code is positioned!

3h1.1K8

Hamza@thegenioo

@ArtificialAnlys @Zai_org Damn, I was expecting only 50 but 1 above is a bonus

But seeing it go way from the efficiency quadrant is kinda not good

3h9425

🆁@peaceinfinitum

@thegenioo @ArtificialAnlys @Zai_org It's a new pareto frontier. It became more efficient.

2h474

leandro@CostaSantos

@ArtificialAnlys @Zai_org @grok can i access glm 5.2 max inside kilo code ?

1h1822

umang payasi@The_neural_post

@ArtificialAnlys @Zai_org @GoogleAIStudio @OpenAI @xai @grok when?

2h5431

Canopy Wave@CanopyWave_AI

@ArtificialAnlys @Zai_org Best open weights model and on the Pareto frontier for cost? This is exactly what the ecosystem needs right now.

2h3324

R1bYamnayaWarrior@R1bYamnayaW

@zephyr_z9 @teortaxesTex Do open-weight LLMs constantly trail closed ones cause they kind of just learn on the closed ones (i.e., open ones will by definition always be behind closed ones), or because they (i assume) operate with less money=less compute?

2h145

Zarko Rashev@rashev_zar8799

@ArtificialAnlys @Zai_org This jump is impressive but it also means that your new updated benchmarks are not accurate for older models. 25% improvement is not realistic for any LLM. So, your scoring system is severely flawed!

3h7212