Latest
  1. Johns Hopkins Bayview tells 3,182 patients an employee put their health data into a public AI tool
  2. China’s AI compute access expected to double in 2027, with US-designed chips about 65 per cent
  3. UK commits £100m to national AI Biosecurity Lab
  4. Moonshot AI releases Kimi K3.1 Pro with 1-million-token context window
  5. Residents in Newport and Cardiff oppose hyperscale datacentres near homes
  6. Study: Grok 4.7 picked Republicans in 92% of forced midterm answers, GPT-6.1 Sol picked Democrats in 97%
  7. Attackers exploiting the PHP flaws hand the break-in to AI agents, Unit 42 says
  8. Elvis Presley’s rightsholders sue Anthropic over training on 274 songs

Latest / Frontier models and safety

Model releaseFrontierChinaConfirmed

Zhipu releases GLM-5, a 744-billion-parameter open model for agentic engineering

See how this story connects

Source: Z.ai: GLM-5

Zhipu AI’s Z.ai released GLM-5 for complex systems engineering and long-horizon agent tasks, scaling from GLM-4.5’s 355 billion parameters (32 billion active) to 744 billion (40 billion active) and from 23 trillion to 28.5 trillion training tokens. It uses DeepSeek Sparse Attention to cut deployment cost and slime, Zhipu’s asynchronous reinforcement-learning system.

Why it matters

A listed Chinese lab shipped a near-trillion-parameter open model that reuses DeepSeek’s attention design, a sign of Chinese labs building on each other’s open work.

When
12 Feb 2026
ID
WTA-N-2026-0190
Where
China
Who
Zhipu AI

Summary by wheresthe.ai, from Z.ai. Checked against the source on 10 Oct 2026.

Read the original at Z.ai

Source copy: download the page as saved on 21 Oct 2026 (PDF). Also held by the Internet Archive.

Next stop on this Storyline: Led toZhipu’s GLM-5.1 tops GPT-5.4 and Opus 4.6 on SWE-Bench Pro, by its own count

Cite this story

WTA-N-2026-0190: “Zhipu releases GLM-5, a 744-billion-parameter open model for agentic engineering.” wheresthe.ai, 12 Feb 2026. https://wheresthe.ai/d/cn-zhipu-glm-5-release-2026-02/ (source: Z.ai: GLM-5).

Spotted an error? Email [email protected].