Zhipu’s New AI Model Takes Aim at OpenAI, Anthropic in China’s Race to Close the Gap
The model adds stronger coding performance and a permissive licence as Z.ai seeks to narrow the gap with OpenAI and Anthropic.
- On Friday, Beijing-based Chinese artificial intelligence company Z.ai launched its GLM-5.3 model, aiming to challenge US leaders with enhanced coding capabilities. The company plans to release the model's weights for public download within two weeks.
- Building on the same 700-billion-parameter base model as June's GLM-5.2, the new iteration focuses on stronger coding and agent performance. Z.ai seeks to outcompete Anthropic and OpenAI by offering accessible, high-performance open-weight models.
- Z.ai reported GLM-5.3 scored 84.5% on CyberGym, slightly surpassing the 83.8% reported for Anthropic's restricted Mythos 5. However, the company said it trailed Mythos 5 on ExploitBench, scoring 54.4% compared to 78.0% in attack-development tests.
- To address safety concerns, the company introduced an "Open Source Shield" initiative to audit projects and added protection layers to screen risky requests. Sensitive cybersecurity functions remain restricted to verified users through a "trusted access" program.
- The launch highlights the intense race between China and the US for AI dominance. By positioning GLM-5.3 as an open-source alternative to restricted models like Mythos, Z.ai aims to broaden access for smaller security teams and developers.
34 Articles
34 Articles
GLM-5.3 is here with advanced cyber capabilities — and reportedly already found a 'serious vulnerability' in Cursor
Chinese AI startup Z.ai, known internationally for its growing lineup of powerful, largely open source GLM series of language models, today released GLM-5.3 with substantial gains in long-horizon coding and a more consequential — and potentially sensitive — jump in cybersecurity capabilities.Already, GLM-5.3's cyber capabilities have found a "potentially serious vulnerability in Cursor," the AI coding startup recently acquired by SpaceX, accordi…
GLM-5.3 didn’t change the base model — where did its coding gains come from?
Z.ai released GLM-5.3 on Friday, a coding and agent model built from the same base model as GLM-5.2. Developers can already use GLM-5.3 through Z.ai’s GLM Coding Plan with Claude Code, Cline, OpenCode and Codex. Direct API access is still listed as “coming soon,” because Z.ai plans to release the model weights after two weeks of hardening and safety testing. Post-training did the heavy lifting Z.ai significantly expanded post-training for GLM-5.…
China's AI Models Are Catching Up. Z.ai's GLM-5.3 Takes Aim at OpenAI & Anthropic
The Beijing-based company said its upcoming GLM-5.3 model will deliver improved coding capabilities, putting its performance closer to leading systems such as Anthropic's Fable 5.
The Chinese artificial intelligence startup Z.ai stated on Friday (14) that its GLM-5.3 open source model came close to the restricted Mythos 5, from Antropic, in identifying software vulnerabilities, strengthening a Chinese competitor who has been gaining adoption among Western developers. Z.ai said that the GLM-5.3 obtained 84.5% score in CyberGym, a test to check if a model can revise code, identify security failures and confirm that they are…
Coverage Details
Bias Distribution
- 47% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
