Would be great if we could add other kinds of AI APIs such as Text to Audio from Eleven Labs and PDF-OCR to MD API from Mistral.