Build guide

How to build a RAG (retrieval-augmented generation) app with Vikasit AI

Retrieval-augmented generation grounds an LLM in your own documents: you embed your content, retrieve the most relevant chunks for a question, and pass them as context. This cuts hallucination and lets the model answer from your data.

Recommended model

Vikasit 3

Strong instruction-following and a low price make it a great default for synthesizing retrieved context into grounded answers. Use Vikasit 3 Max for harder, multi-document reasoning.

Steps

1
Chunk your documents (e.g. 500–1000 tokens with overlap) and store them.
2
Embed each chunk and store the vectors in a vector database (FAISS, pgvector, Qdrant, etc.).
3
At query time, embed the user question and retrieve the top-k most similar chunks.
4
Build a prompt that includes the retrieved chunks as context plus the question.
5
Call the Vikasit chat API and instruct the model to answer only from the provided context.
6
Return the answer along with citations to the source chunks for trust.

Code

The Vikasit Inference API is OpenAI-compatible, so this uses the standard OpenAI Python SDK pointed at https://api.vikasit.ai/v1.

rag-app.py

from openai import OpenAI

client = OpenAI(
    base_url="https://api.vikasit.ai/v1",
    api_key="sk-vikasit-...",  # get one at vikasit.ai/auth
)

def answer(question: str, retrieved_chunks: list[str]) -> str:
    context = "\n\n".join(retrieved_chunks)
    resp = client.chat.completions.create(
        model="vikasit-3",
        messages=[
            {
                "role": "system",
                "content": (
                    "Answer using ONLY the context below. "
                    "If the answer isn't there, say you don't know.\n\n"
                    f"Context:\n{context}"
                ),
            },
            {"role": "user", "content": question},
        ],
    )
    return resp.choices[0].message.content

Build your RAG (retrieval-augmented generation) app today

Get an API key and 2M free tokens a day on Vikasit Nova. Pay-as-you-go, no minimums, OpenAI-compatible.

Get an API key See all models & pricing