Skip to main content
Optimise how documents are split into chunks for better retrieval quality. 
from praisonaiagents import Agent, KnowledgeConfig

agent = Agent(
    name="Researcher",
    instructions="Answer using retrieved document chunks.",
    knowledge=KnowledgeConfig(sources=["docs/"], chunk_size=500, chunk_overlap=50),
)

agent.start("What does our deployment guide say about Docker?")
The user uploads documents; they are chunked and embedded, and the agent retrieves the best slices for each question. Optimize how documents are split into chunks for better retrieval quality.

​
Chunking Options

from praisonaiagents import Knowledge

knowledge = Knowledge(
    sources=["docs/"],
    chunk_size=500,        # Target chunk size
    chunk_overlap=50,      # Overlap between chunks
    chunking_strategy="semantic"  # Chunking method
)

​
Strategies

​
Fixed Size (Default)

knowledge = Knowledge(
    sources=["docs/"],
    chunking_strategy="fixed",
    chunk_size=500,
    chunk_overlap=50
)

​
Semantic

knowledge = Knowledge(
    sources=["docs/"],
    chunking_strategy="semantic"
)

​
Sentence-Based

knowledge = Knowledge(
    sources=["docs/"],
    chunking_strategy="sentence",
    sentences_per_chunk=5
)

​
Best Practices

Technical docs: 500–800 tokens with 50–100 overlap (semantic). FAQs: 200–400 with 20–50 overlap (fixed).
Use 800–1200 token chunks with 100–200 overlap and semantic splitting for narrative content.
Keep code chunks around 300–500 tokens with ~50 overlap; prefer fixed splitting for source files.

Knowledge Base

Build a knowledge base

Knowledge Module

Full API reference