Prompting & Structured Output · Beginner
Context Window Chunking
See how a long prompt is split into chunks to fit each model's context window.
Context windows
Live chunk split
Token budget
Understanding Text Chunking
Context Window
The maximum tokens an AI model can process at once, including both input and output.
Different models have different window sizes
Small: 4K–16K · Mid: 8K–128K · Long-context: 200K–1M+ tokens
Chunking Strategy
Breaking large text into smaller, manageable pieces that fit within context limits.
Each chunk maintains context and meaning
Overlapping chunks preserve continuity
Context Overflow
When text exceeds the model's limits, it must be truncated or split to avoid errors.
Red chunks indicate potential overflow
Plan chunking to avoid losing important content
Chunking Best Practices
Semantic Boundaries: Split at natural breaks like paragraphs or sections
Overlap Strategy: Include 10-20% overlap between chunks for context preservation
Size Optimization: Balance chunk size with processing efficiency and context retention
Model Selection: Choose models with appropriate context windows for your use case
Long Prompt Input
0
Total Tokens
0
Chunks Needed
0
Context Window
0%
Context Usage
Small (4K)
4,096 tokens
0
Mid-size (8K)
8,192 tokens
0
Long-context (100K)
100,000 tokens
0
Chunked Output
Normal chunk (fits in context)
Overflow chunk (exceeds context)
Enter a long prompt above to see how it gets chunked for different models