# Z.ai Launches GLM-5.3 With Frontier Coding and Unexpected Cybersecurity Capability

> Chinese lab Z.ai has released GLM-5.3, and the notable part is how it got better. The team kept the exact same 743-billion-parameter base model as GLM-5.2 and pushed all the gains through additional post-training rather than an expensive from-scratch retrain. The payoff is large. Its score on the Terminal-Bench 3.0 agentic coding test leapt from 4.6 to 28.3, a roughly sixfold jump on long-horizon tasks that require a model to plan and execute across many steps. The stranger result showed up in security. The same post-training that sharpened coding also taught the model to reason through exploit chains, and in testing it surfaced 1,097 critical bugs across Linux, WebKit, and FreeBSD. On the CyberGym benchmark it reached 84.5%, edging past the frontier scores Z.ai cites for Anthropic and OpenAI's latest systems. That dual-use edge is exactly why this matters. A model that finds real vulnerabilities is a gift to defenders and a hazard in the wrong hands, and Z.ai says it is holding open weights back for about two weeks pending a safety review. For now it is available through the company's API and coding plan.

_Section: [Daily AI Updates](https://www.wortins.com/daily-ai) · Source: MarkTechPost · Published Saturday, August 15, 2026_

## Wortins' read

Chinese lab Z.ai has released GLM-5.3, and the notable part is how it got better. The team kept the exact same 743-billion-parameter base model as GLM-5.2 and pushed all the gains through additional post-training rather than an expensive from-scratch retrain. The payoff is large. Its score on the Terminal-Bench 3.0 agentic coding test leapt from 4.6 to 28.3, a roughly sixfold jump on long-horizon tasks that require a model to plan and execute across many steps. The stranger result showed up in security. The same post-training that sharpened coding also taught the model to reason through exploit chains, and in testing it surfaced 1,097 critical bugs across Linux, WebKit, and FreeBSD. On the CyberGym benchmark it reached 84.5%, edging past the frontier scores Z.ai cites for Anthropic and OpenAI's latest systems. That dual-use edge is exactly why this matters. A model that finds real vulnerabilities is a gift to defenders and a hazard in the wrong hands, and Z.ai says it is holding open weights back for about two weeks pending a safety review. For now it is available through the company's API and coding plan.

## Source

[Read the full story at MarkTechPost](https://www.marktechpost.com/2026/08/14/z-ai-ships-glm-5-3-without-retraining-the-base-model-better-at-complex-coding-and-long-horizon-tasks/)

## Related coverage

- [80% of Organizations Report Measurable ROI from AI Agents in Production](https://www.wortins.com/story/80-of-organizations-report-measurable-roi-from-ai-agents-in--0f1c628d) — [Anthropic](https://resources.anthropic.com/2026-state-of-ai-agents)
- [SoftBank Plans Record $6.3 Billion Retail Bond Sale to Fund OpenAI Investment](https://www.wortins.com/story/softbank-plans-record-6-3-billion-retail-bond-sale-to-fund-o-dbc93b82) — [Bloomberg](https://www.bloomberg.com/news/videos/2026-08-20/bloomberg-tech-8-20-2026-video)
- [Cohere Launches Command A+ Mixture-of-Experts Model](https://www.wortins.com/story/cohere-launches-command-a-mixture-of-experts-model-5d840970) — [Cohere](https://docs.cohere.com/docs/command-a-plus)
- [Stripe Acquires OpenRouter for $7B+](https://www.wortins.com/story/stripe-acquires-openrouter-for-7b-2abe2a10) — [TechCrunch](https://techcrunch.com/2026/08/16/stripe-will-reportedly-acquire-ai-gateway-startup-openrouter-for-7b/)
- [Chinese AI Models Now 60% of OpenRouter Traffic, Surpassing US Market Share](https://www.wortins.com/story/chinese-ai-models-now-60-of-openrouter-traffic-surpassing-us-ff49998d) — [Fortune](https://fortune.com/2026/08/21/what-is-ai-death-zone-china-models-open-source/)
- [DARPA and US Air Force Successfully Fly F-16 Fighter Jet Under Full AI Control](https://www.wortins.com/story/darpa-and-us-air-force-successfully-fly-f-16-fighter-jet-und-b659da37) — [DARPA](https://www.darpa.mil/news/2026/darpa-us-air-force-fly-ai-controlled-f-16)

---

_Curated and written by [Wortins](https://www.wortins.com) — The daily AI briefing. Every story links to its original source; the "Wortins read" on each is our own original analysis. [About Wortins & our editorial approach](https://www.wortins.com/about)._