Octonous runs on a default AI model. If you have a specific model you'd prefer (your own deployment of Claude or GPT-4, for example), you can connect it.
How to add a custom model:
Go to Settings → Custom models
Click Add model
Enter the API endpoint and your API key
Give it a name and save
Once added, select it when starting a conversation or building an automation.
Why this is useful:
Use a model you've already paid for to avoid double costs
Use a fine-tuned or internally hosted model
Test different models for specific tasks
Your API key is encrypted at rest using KMS envelope encryption.