This feature introduces a bring your own LLM integration - not only in the .env but also in the Account settings.
Users will be able to select their provider and model from the Account settings in the UI. They can also do this in the .env.
First, we need to enhance the .env features. Currently, only OpenAI is supported. I need support for Mistral, Claude, OpenRouter, Gemini. This all works via API keys. Except for Google Vertex. Vertex support needs to work also through the gcloud authentication method (next to the API method).
In the UI, under Account Settings allow users to select a specific provider and paste their credentials into it as well. The sensitive information is handled securely. Just like the other integrations. And users get the option to select specific models in the Account Settings.
This feature introduces a bring your own LLM integration - not only in the .env but also in the Account settings.
Users will be able to select their provider and model from the Account settings in the UI. They can also do this in the .env.
First, we need to enhance the .env features. Currently, only OpenAI is supported. I need support for Mistral, Claude, OpenRouter, Gemini. This all works via API keys. Except for Google Vertex. Vertex support needs to work also through the gcloud authentication method (next to the API method).
In the UI, under Account Settings allow users to select a specific provider and paste their credentials into it as well. The sensitive information is handled securely. Just like the other integrations. And users get the option to select specific models in the Account Settings.