Knowledge Base Article

AI Settings: General

 

AI Settings > General: enable AI, choose Primary/Fallback LLM, and Primary/Fallback Embedding models.

Enable AI Settings

Turn on the master switch to enable AI features for your community. If this is off, CLM and Answer Assist will not run.

AI Models: Primary and Fallback

Your Primary model is used for tasks like content summarization, question validation, and answer generation. Your Fallback model is used if the Primary model is temporarily unavailable.

Default models (recommended)
Your instance is already pre-configured with recommended defaults:
• Primary LLM: OpenAI GPT 5 Nano
• Fallback LLM: OpenAI GPT 4o Mini
You can change models if needed, but you do not need to change them to run accurate tests.

Embedding Models: Primary and Fallback

Embedding models generate vector embeddings that power semantic search. This is what allows CLM/AA to find relevant content even when keywords do not match exactly.

Default embedding model (recommended)
• Primary Embeddings: OpenAI Text Embedding 3 Small
• Fallback Embeddings: OpenAI Text Embedding 3 Small
Keeping Primary and Fallback embeddings consistent helps you produce stable search behavior during initial setup.

Other than the default models, there are many others that are supported, plus you also have the option to bring your own API key.

Supported AI models
Updated 3 months ago
Version 4.0

1 Comment

  • allensmith81's avatar
    allensmith81
    Boss
    2 days ago

    I have tried to bring my own model but nothing I do will work, it tried setting up the endpoints, i put the api key in the API key field but it would hit the AI and get a 401 error. I tried removing the API Key and putting it in the header as {"api-key":"MYKEY"} still got a 401 error.

    Do you have more documentation on using your Own Model?