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

# Hybrid Search

> Combine vector similarity with BM25 keyword search for better retrieval.

Hybrid search combines vector similarity (semantic) with BM25 keyword search (exact) for more accurate retrieval.

<Steps>
  <Step title="Create the knowledge base">
    ```python hybrid_search.py theme={null}
    import asyncio
    from definable.agent import Agent
    from definable.knowledge import Knowledge, FTSIndex, HybridSearchConfig
    from definable.vectordb import InMemoryVectorDB

    # Create FTS index (must initialize before use)
    fts = FTSIndex()

    async def main():
        await fts.initialize()

        knowledge = Knowledge(
            vector_db=InMemoryVectorDB(),
            fts_index=fts,
            hybrid_config=HybridSearchConfig(
                merge_strategy="rrf",  # reciprocal rank fusion
                vector_weight=0.7,
                text_weight=0.3,
            ),
        )

        # Add documents (auto-indexed in both vector DB and FTS)
        await knowledge.aadd("Python 3.12 introduced new syntax features.")
        await knowledge.aadd("Definable requires Python 3.12 or higher.")
        await knowledge.aadd("The framework uses async/await for non-blocking I/O.")

        agent = Agent(
            model="gpt-4o",
            knowledge=knowledge,
            instructions="Answer using the knowledge base.",
        )

        output = await agent.arun("What Python version does Definable need?")
        print(output.content)

    asyncio.run(main())
    ```
  </Step>

  <Step title="Install and run">
    <Snippet file="install-definable.mdx" />

    ```bash theme={null}
    python hybrid_search.py
    ```
  </Step>
</Steps>

<Warning>
  `FTSIndex` must be initialized with `await fts.initialize()` before any search or add operations.
</Warning>