China's Zhipu AI has achieved a notable milestone in the AI world. Its latest model, GLM-5.2, secured the top position in a respected creative coding benchmark.
On June 20, 2026, Design Arena announced on X that GLM-5.2 reached #1 in its single-round HTML web design leaderboard for the non-agent category.
>>> GPT-5.6 Rumors Intensify as OpenAI Reportedly Prepares Major AI Upgrade
The platform is the first large crowdsourced benchmark where real people blindly vote on AI-generated designs.
GLM-5.2 outperformed Claude Fable 5, as well as Opus 4.6 and 4.7 versions.
It jumped five places from its predecessor, GLM-5.1, achieving an Elo score of around 1360 in the Code Categories arena.
The model excels at clean layouts, integrating images from CDNs, typography, visual hierarchy, and subtle animations. It also works well with libraries like Chart.
js and Three. js.
>>> OnePlus N6 Geekbench Listing Confirms Dimensity 6300 Chipset Ahead of June 30 Launch
Its win rate improved by about 6 percentage points this time.
GLM-5.2 relies heavily on Tailwind CSS, appearing in 91% of its designs, and Font Awesome in 51%.
In contrast, Fable 5 used Tailwind in only about 57% of its sessions.
Pricing is another advantage.
>>> Nia Sioux Turns 25: See Her Stunning Transformation from 'Dance Moms' Star
GLM-5.2's API costs around $1.40 per million input tokens and $4.40 for output, significantly cheaper than Fable 5's $10 and $50.
It is an open-weights model under the MIT license with a 1M-token context window, allowing local deployment and longer projects.
Design Arena's results are based on millions of votes from creators focusing on real aesthetics and usability.
This win places GLM-5.2 among the strongest open models in broader coding benchmarks.
The achievement signals that Chinese labs are advancing rapidly in specialized creative areas.
>>> Sombr Scores Invite to Taylor Swift and Travis Kelce's Wedding Celebration
GLM-5.2's combination of strong performance, low cost, and openness may make it a preferred choice for many web development teams.
