Name of the LLM, such as "gpt-3.5-turbo"
The pricing model for the embedding service, if known. Null for local models (Ollama, ONNX) where the cost is zero.
Name of the provider, such as "OpenAI"