US EditionPublished • loading... • Updated
Z.ai Releases GLM-4.7 Designed for Real-World Development Environments, Cementing Itself as "China's OpenAI"
GLM-4.7 improves multi-step task handling and coding stability in real-world environments, achieving top scores on benchmarks including an 87.4 on IJ-Bench, the highest for open-source models.
- On December 22, Z.ai released GLM-4.7 in Beijing, making it available via the BigModel.cn API and setting it as the default model for the GLM Coding Plan.
- Built on GLM-4.6, GLM-4.7 focuses on supporting developers and development teams with multi-step workflows, training and evaluation designed for real-world constraints.
- Benchmark testing shows GLM-4.7 outperforms GLM-4.6 and matches Claude Sonnet 4.5, scoring 87.4 on IJ-Bench, 67.5 on BrowseComp, and tested on 100 real programming tasks in Claude Code.
- Developers and partners have already integrated the GLM Coding Plan into TRAE, Cerebras, YouWare and Vercel, and Z.ai says it aims to list on the Stock Exchange of Hong Kong.
- The company reported 57.4 million RMB, 124.5 million RMB, and 312.4 million RMB in revenue from 2022–2024, with 190 million RMB in H1 2025, supporting rapid growth, Z.ai says.
Insights by Ground AI
29 Articles
29 Articles
The Daily Advance+28 Reposted by 28 other sources
Z.ai Releases GLM-4.7 Designed for Real-World Development Environments, Cementing Itself as "China's OpenAI"
BEIJING, Dec. 27, 2025 /PRNewswire/ -- On December 22, Z.ai released GLM-4.7, the latest iteration of its GLM large language model family. Designed to handle multi-step tasks in production, GLM-4.7 targets development environments that involve lengthy task cycles, frequent tool…
Coverage Details
Total News Sources29
Leaning Left5Leaning Right3Center7Last UpdatedBias Distribution47% Center
Bias Distribution
- 47% of the sources are Center
47% Center
L 33%
C 47%
R 20%
Factuality
To view factuality data please Upgrade to Premium
