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

# Chunking

> Split documents into optimal chunks for RAG

Chunking splits documents into pieces for better retrieval.

```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
graph LR
    subgraph "Chunking"
        D[📄 Document] --> C[✂️ Chunker]
        C --> C1[📝 Chunk 1]
        C --> C2[📝 Chunk 2]
        C --> C3[📝 Chunk 3]
    end
    
    classDef doc fill:#6366F1,stroke:#7C90A0,color:#fff
    classDef chunk fill:#10B981,stroke:#7C90A0,color:#fff
    
    class D doc
    class C,C1,C2,C3 chunk
```

## Quick Start

<Steps>
  <Step title="Create Agent with Chunked Knowledge">
    ```rust theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
    use praisonai::{Agent, Knowledge, ChunkingConfig};

    // Configure chunking for knowledge base
    let knowledge = Knowledge::new()
        .source("docs/")
        .chunking(ChunkingConfig::semantic()
            .chunk_size(1000)
            .overlap(200))
        .build()?;

    // Add documents - they'll be automatically chunked
    knowledge.add_file("guide.pdf").await?;

    // Create agent that uses chunked knowledge
    let agent = Agent::new()
        .name("Assistant")
        .instructions("You answer questions using the knowledge base")
        .build()?;

    // Search returns relevant chunks
    let chunks = knowledge.search("How do I get started?").await?;
    ```
  </Step>
</Steps>

***

## Chunking Strategies

```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
graph TB
    subgraph "Choose Strategy"
        Q{Content Type?} --> |Technical| F[Fixed - 500 chars]
        Q --> |Narrative| S[Semantic - by meaning]
        Q --> |Structured| P[Paragraph - natural breaks]
    end
    
    classDef question fill:#6366F1,stroke:#7C90A0,color:#fff
    classDef strategy fill:#10B981,stroke:#7C90A0,color:#fff
    
    class Q question
    class F,S,P strategy
```

| Strategy    | Best For             |
| ----------- | -------------------- |
| `Fixed`     | Code, technical docs |
| `Semantic`  | Articles, narratives |
| `Sentence`  | Q\&A, FAQs           |
| `Paragraph` | Structured documents |

***

## Configuration

| Option          | Type               | Default    | Description            |
| --------------- | ------------------ | ---------- | ---------------------- |
| `chunk_size`    | `usize`            | `1000`     | Characters per chunk   |
| `chunk_overlap` | `usize`            | `200`      | Overlap between chunks |
| `strategy`      | `ChunkingStrategy` | `Semantic` | How to split           |

***

## Related

<CardGroup cols={2}>
  <Card title="Knowledge" icon="book" href="/docs/rust/knowledge">
    RAG system
  </Card>

  <Card title="Embeddings" icon="vector-square" href="/docs/rust/embeddings">
    Vector embeddings
  </Card>
</CardGroup>