I am currently developing an AI-powered risk assessment and adverse media screening dashboard that integrates the Perplexity API to analyze financial crime-related news and entities. The project is designed to demonstrate real-time due diligence, risk scoring, and risk visualization using Perplexity’s large language models.
However, I am encountering repeated errors when attempting to invoke the API. The response consistently shows:
{“error”: {“message”: “Invalid model ‘pplx-7b-chat’. Permitted models can be found in the documentation at https://docs.perplexity.ai/getting-started/models.”, “type”: “invalid_model”, “code”: 400}}
I have tried multiple model names, including "pplx-7b-chat", "pplx-70b-chat", and "pplx-search-v1", but all return the same invalid model message. Additionally, attempting to access the /models endpoint results in a 404 error.
Could you please assist with the following:
-
Confirm if my current API key includes permission for model access.
-
Provide the correct model names that are accessible for my API key. I have Perplexity Pro.
For reference, my API key follows the format:
pplx-****
Technical Summary:
The project runs in both Google Colab and Python environments.
It utilizes a lightweight analyzer script that sends dynamic queries to the Perplexity API and presents risk scores through a dark-mode HTML dashboard. The dashboard visually ranks entities as Low, Medium, or High risk, based on detected fraud, money laundering, or sanction terms within Perplexity’s search responses.
I am preparing this as part of my AI compliance presentation, so timely activation or guidance would be greatly appreciated.