The Chat Completion action works by sending a request to system running the large language model.

These systems offer an API that listens on published 'Endpoints'. Each system may offer several endpoints, one being the 'Chat Completion' endpoint.

The Chat Completion action will work with any Chat Completion endpoint that is compatible with the OpenAI API endpoint.

Public Endpoints

There are many services providing endpoints. To use these you normally need to open an account with them, and obtain an API Key that identifies who you are, and where they are charging, to allow them to bill for usage.

New services are constantly being launched, so this is not a comprehensive list, but some common ones are:

OpenAI

https://api.openai.com/v1/chat/completions

Google Gemini

https://generativelanguage.googleapis.com/v1beta/openai/chat/completions/

Deepseek

https://api.deepseek.com/chat/completions

Mistral

https://api.mistral.ai/v1/chat/completions

Private Endpoints

If you run your own service then it will provide an endpoint, which will be the IP address or domain name of the server where it installed. If you are running this on the same server as clearString (not recommended for performance reasons, but useful when testing) then you can use localhost instead of the IP address or domain name. Often these services will default to providing the API on a non-standard port.

Ollama

This provides a service on port 11434.

http://localhost:11434/v1/chat/completions

LM Studio

This provides a service on port 1234.

http://localhost:1234/v1/chat/completions