> ## Documentation Index
> Fetch the complete documentation index at: https://docs.agno.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Traditional RAG with PgVector

> Retrieve from PgVector for a string input and append references before the first model call.

For a string-input run, set `add_knowledge_to_context=True` to retrieve before the first model call.

```python traditional_rag_pgvector.py theme={null}
from agno.agent import Agent
from agno.knowledge.embedder.openai import OpenAIEmbedder
from agno.knowledge.knowledge import Knowledge
from agno.models.openai import OpenAIResponses
from agno.vectordb.pgvector import PgVector, SearchType

db_url = "postgresql+psycopg://ai:ai@localhost:5532/ai"

knowledge = Knowledge(
    vector_db=PgVector(
        table_name="recipes",
        db_url=db_url,
        search_type=SearchType.hybrid,
        embedder=OpenAIEmbedder(id="text-embedding-3-small"),
    ),
)

agent = Agent(
    model=OpenAIResponses(id="gpt-5.2"),
    knowledge=knowledge,
    add_knowledge_to_context=True,
    search_knowledge=False,
    markdown=True,
)

if __name__ == "__main__":
    knowledge.insert(
        name="Thai Recipes",
        url="https://agno-public.s3.amazonaws.com/recipes/ThaiRecipes.pdf",
    )
    agent.print_response(
        "How do I make chicken and galangal in coconut milk soup?",
        stream=True,
    )
```

## Run the Agent

<Steps>
  <Snippet file="run-pgvector-step.mdx" />

  <Snippet file="create-venv-step.mdx" />

  <Step title="Install dependencies">
    ```bash theme={null}
    uv pip install -U agno openai pgvector psycopg pypdf sqlalchemy
    ```
  </Step>

  <Step title="Export the API key">
    ```bash theme={null}
    export OPENAI_API_KEY=your_openai_api_key_here
    ```
  </Step>

  <Step title="Run the agent">
    ```bash theme={null}
    python traditional_rag_pgvector.py
    ```
  </Step>
</Steps>

## How It Works

1. `add_knowledge_to_context=True` searches with the run's string input.
2. Returned documents are appended to the user message inside a `<references>` block.
3. `search_knowledge=False` removes the model-callable search tool.

## Next Steps

| Task                                | Guide                                                                   |
| ----------------------------------- | ----------------------------------------------------------------------- |
| Let the model choose when to search | [Agentic RAG with PgVector](/knowledge/agents/agentic-rag-pgvector)     |
| Tune hybrid ranking                 | [Hybrid Search](/knowledge/concepts/search-and-retrieval/hybrid-search) |
| Apply metadata filters              | [Filtering](/knowledge/concepts/filters/overview)                       |