[KongchangAI]
Model默认 Neon High

Neon High

The default thinking-level configuration for the Gemini 4 Neon model, compressing the thinking budget to control latency; it performs less well than the high thinking level on complex tasks.

Source Articles