Tech CEOs hail China’s GLM-5.2 as top open-weights AI model

Chinese Open-Source GLM-5.2 Tops All Google Models on AI Index

Z.AI’s GLM-5.2, released on June 13, 2026, has vaulted to the top of the open-weights category on the Artificial Analysis Intelligence Index, scoring 51 on version 4.1 of the benchmark and placing fourth overall globally — ahead of every Google model currently available. The 744-billion-parameter Mixture-of-Experts model, released under a permissive MIT license with weights freely downloadable from Hugging Face, has drawn immediate praise from prominent tech leaders and reignited debate about the competitive dynamics between Chinese and Western AI labs.ascii

Benchmark Performance

GLM-5.2’s score of 51 on the Artificial Analysis Intelligence Index places it behind only Claude Fable 5, Claude Opus 4.8, and OpenAI’s GPT-5.5 at high reasoning effort — and well ahead of Google’s Gemini 3.1 Pro Preview, which scores 46. Among open-weights models, it leads MiniMax-M3 and DeepSeek V4 Pro Max, both at 44, by a seven-point margin.hyper

On coding benchmarks, the model scored 62.1 on SWE-bench Pro, surpassing GPT-5.5’s 58.6, and 81.0 on Terminal-Bench 2.1. It uses 40 billion active parameters per inference call despite its 744 billion total parameters, with gains over its predecessor GLM-5.1 coming entirely from training improvements rather than scale.officechai

Industry Reaction

Guillermo Rauch, CEO of Vercel, said he was “genuinely impressed, almost shocked” by GLM-5.2’s coding ability, adding: “This changes things.” Box CEO Aaron Levie called the state of open-weights AI “pretty remarkable,” noting that as the gap between open and closed models remains narrow, more value can be built on the applied layer. Jeremy Howard, co-founder of Answer.AI and Fast.AI, called the model “a marvel” and placed it on par with Claude Opus 4.8 and GPT-5.5.aiproductivity

Strategic Context

The model ships at $1.40 per million input tokens — a fraction of what closed frontier models charge — with a 1-million-token context window expanded from GLM-5.1’s 200,000 tokens. Z.AI, formerly known as Zhipu AI, completed a Hong Kong IPO in January 2026 and has maintained its development pipeline without any Nvidia hardware since being placed on the US Entity List.note

The release echoes the disruption caused by DeepSeek in early 2025, which forced a reassessment of Chinese AI capabilities. As one analysis noted, Google’s Gemini team has “been struggling as of late,” with its models failing to keep pace with both Chinese open-source labs and Western closed-model competitors.vectorlab