Custom deployed model usage

[uttambarthakur]

Hi,

Does extract thinker expose APIs to use custom model. My models are deployed in a secured environment. I want to use those inference APIs instead of directly using the OpenAI or Grok APIs.
How can I set up the connection string and credentials to use such models?

[enoch3712]

Yes ofc! You just need to pass the credentials and pass the right string to map for your model:

https://enoch3712.github.io/ExtractThinker/examples/local-processing/

You can more at Litellm, that is the underground LLM agnostic model:

https://docs.litellm.ai/docs/