BREAKING
GLM-5.2 Hits No. 2 on Code Arena
Frontend Leaderboard Rankings
Overall
2
React
2
HTML
4
0
Terminal-Bench 2.1
0
SWE-bench Pro
0
%
win vs Opus 4.7
0
$
input /M tokens
0
$
output /M tokens
0
M
context window
GLM-5.2 vs Claude Opus 4.8
GLM-5.2 (Max)
MIT
●
744B params, ~40B active
●
$1.40 / $4.40 per M
●
Strong frontend design taste
Claude Opus 4.8
closed
●
85.0 Terminal-Bench
●
$5 / $25 per M
●
Tighter instruction following
Open Weights Close the Gap
AI NEWS BLITZ
Z.ai's open-weight GLM-5.2 just landed at number two on Code Arena's frontend coding board.