RAG-Buddy API Reference and Playground

Welcome to the RAG-Buddy API Reference and Playground. This interactive tool is designed to help developers understand and test the RAG-Buddy API endpoints.

The playground API goes through an extra server hop which might result in a corrupted response. This seems especially the case for error paths. We are working with the playground provider to get this fixed. In the meantime, we recommend using the API directly. Simply copy the generated example code in curl or you favorite programming language and run it from your local machine.

API Endpoints

The RAG-Buddy API provides the following endpoints:

  1. Text Classification (/tc/{version}): This endpoint is used for the Text-Classification use case. Try it here: /tc playground.

  2. Rag Citation (/ragc/{version}): This endpoint is used for the RAG-Citation use case. Try it here: /ragc playground.

  3. Proxy (/proxy/{cache_type}/{version}/{path}): This endpoint is used to transparently read and write to cache and forward requests to the LLM API. Try it here: /proxy playground.

The OpenAI API might return compressed responses, regardless of the request headers. Most of the generated sample code handles this by automatically decompressing the response body, but this is not the case for curl. Add the --compressed flag to the curl command to fix this.


To authenticate, you need to include the helvia-rag-buddy-token header in your request headers. This token is required for all API endpoints and can be generated in your project on the RAG-Buddy console.

Additionally, the playground requires an OpenAI API key. This key should be sent as a Bearer token in the Authorization header. Here’s an example of how to include it in your requests:

Authorization: Bearer YOUR_OPENAI_API_KEY

Replace YOUR_OPENAI_API_KEY with your actual OpenAI API key.

Need assistance?

Please contact the RAG-Buddy Support Team at ragbuddy@helvia.ai for any issues or further information.