modelgrep

Qwen: Qwen3 30B A3B Instruct 2507 vs Z.ai: GLM 4.5V

Qwen: Qwen3 30B A3B Instruct 2507 wins on more metrics (6 of 8), but the right pick depends on what you optimize for — see the breakdown below.

MetricQwen: Qwen3 30B A3B Instruct 2507Z.ai: GLM 4.5V
Intelligence Index15.015.1
Coding Index14.210.9
GPQA Diamond66%68%
Design Arena Elo
Speed (tokens/sec)4519
Latency245ms1.0s
Input price /M$0.048$0.600
Output price /M$0.193$1.80
Context window131K66K
CapabilitiesToolsJSONReasoningToolsJSONVision