This is an openAI compatible client library for text-generation-webui.
For an example see api_test which showcases the low level, high level and node API. I generally recommend just using the TextgenChatNode as this is the simplest to use and uses a shared connection through a TextgenServer singleton. This is also written asynchronously so it should be a lot easier to use.
If you don't use the default config you'll need to configure the port and server address in core/textgen_api_connector.gd. Everything is based on this openAI client (the low level API). The server can be configured by updating the singleton scene in servers/textgen_server.tscn.
This project is licensed under the AGPLv3 since I just copied over the API configs from the webui which is licensed under AGPLv3. I plan to update the files under config/ and remove the internal API in core/textgen_api_connector.gd in the future so this can be licensed under MIT under another repo. This project itself is adapted from a very old OpenAI client I wrote based on the openAI python library so there'll be no licensing issues there. I'll either target vLLM or llama.cpp, it should even work now, but never tested with another API.