modelgrep

Anthropic: Claude Sonnet 4.6 vs Z.ai: GLM 5.2

Z.ai: GLM 5.2 wins on more metrics (8 of 9), but the right pick depends on what you optimize for — see the breakdown below.

MetricAnthropic: Claude Sonnet 4.6Z.ai: GLM 5.2
Intelligence Index51.1
Coding Index43.068.8
GPQA Diamond80%90%
Design Arena Elo13261369
Speed (tokens/sec)5458
Latency1.0s1.2s
Input price /M$3.00$1.20
Output price /M$15.00$4.10
Context window1M1.0M
CapabilitiesReasoningToolsJSONVisionReasoningToolsJSON