model selection in OpenAIRealTimeModel and GoogleGeminiLiveMode

[stc1988]

I made the following changes in relation to #1547 (comment).

• model selection in GoogleGeminiLiveModel
  • Set the latest model gemini-2.5-flash-native-audio-preview-12-2025 as the default model.
• model selection in OpenAIRealTimeModel
  • Set a more cost-effective and faster model gpt-realtime-mini as the default model.

[phoddie]

I took a quick look. This seems sufficient for the API to be able to choose the model.

t feels a little inconsistent in conversationalAI because the UI cannot select the models for Google and OpenAI, like they can for Hume or ElevenLabs. Maybe it makes sense to fill in the providers section for OpenAI and Google as part of this change. What do you think?

[stc1988]

Unlike others, OpenAI and Google will likely require selection from their own real-time AI models.
Therefore, the model options may be limited, but I will consider it for consistency with others.

[phoddie]

Unlike others, OpenAI and Google will likely require selection from their own real-time AI models.

I expect that.

[stc1988]

I added providers section for OpenAI and Google.

[phoddie]

Thank you. We are reviewing the changes.