Model updates take too long

I don’t understand why it often takes several weeks or even months until new models become available here. For instance, Gemini 3.5 Flash would offer a fantastic balance between speed, intelligence and cost. GPT 5.5 has been released 5 weeks ago. For Opus your supported version is 4 months behind the curve. And since quite some time the frontier models come with a set of controls such as “effort” or “fast mode” that could change things for customer support intelligence.

I’m not saying we need every single update instantly. But I can’t see a clear strategy behind the model selection and how and when updates come to the platform.