Service URL

... is the URL of the actual AI service API.

The service URL can be specified as part of the Service Properties. If it's not specified then the service-specific default URL is used.

The final URL will then be caculated by:

adding the path of the service-specific default URL to the service URL
- to skip adding the default path, add the anker #no-default-path to the service URL
replacing the ${engine} placeholder (if used in the URL) with the actual engine name.

The default URLs (and more) can be found in the Services list.

Examples

Service URL (optional)	Engine	Default URL	Final URL used
-	gpt-3.5	https://api.openai.com/v1/chat/completions	https://api.openai.com/v1/chat/completions
https://openai-compatible-api.example.com/	gpt-3.5	https://api.openai.com/v1/chat/completions	https://openai-compatible-api.example.com/v1/chat/completions
https://openai-compatible-api.example.com/api/chat-completions#no-default-path	gpt-3.5	https://api.openai.com/v1/chat/completions	https://openai-compatible-api.example.com/api/chat-completions

-	microsoft/DialoGPT-large	https://api-inference.huggingface.co/models/${engine}	https://api-inference.huggingface.co/models/microsoft/DialoGPT-large
https://my-hf-inference.example.com/	microsoft/DialoGPT-large	https://api-inference.huggingface.co/models/${engine}	https://my-hf-inference.example.com/models/microsoft/DialoGPT-large
https://my-hf-inference.example.com/api-start/v1/	microsoft/DialoGPT-large	https://api-inference.huggingface.co/models/${engine}	https://my-hf-inference.example.com/api-start/v1/models/microsoft/DialoGPT-large
https://my-hf-inference.example.com/api-m/${engine}#no-default-path	microsoft/DialoGPT-large	https://api-inference.huggingface.co/models/${engine}	https://my-hf-inference.example.com/api-m/microsoft/DialoGPT-large
https://my-hf-inference.example.com/api-models/foo#no-default-path	microsoft/DialoGPT-large	https://api-inference.huggingface.co/models/${engine}	https://my-hf-inference.example.com/api-models/foo