[Feature Request] Improved/streamlined API Key management and model selection

I want to see one of the two options please:

Option 1: API On/Off Switch with corresponding model availability

  • Request: Add a toggle switch to easily enable or disable API keys for specific models.
  • Benefits: Streamlined workflow, reduced frustration, and easier model selection.

Option 2: Unified model access (propietary + API)

  • Request: Allow users to access all models, including API models, with a setting that acknowledges the user’s understanding of the risks and limitations.
  • Benefits: Unified access to all models, increased flexibility and control, and user accountability.

By implementing on of these two components, users will have a more seamless and efficient experience when working with different models.

Edit: Also, a toggle to choose if model choice stays consistent or not for each CMD+K, Chat, Interpreter and 200K mode would also be great. (e.g. Haiku/LLaMA-70B for CMD+K, Opus for chat, GPT-4 for Interpreter mode in a fixed manner unless it’s explicitly changed in that mode)


The continue-dev model/provider config is pretty great, when I’m switching provider at least I don’t need to get the API key again. Would like the same


Something like this indeed:

1 Like