Markdown Chunk Splitter

Set chunk size and overlap to split long Markdown or plain text. Preview RAG chunking before embedding and retrieval tuning.

Read the full guide: LLM Prompts and Context: Token Budgets and Templates →

Privacy: processed locally, never uploaded.

↓ Paste in the input area below to see results instantly

Split long documents by character count with optional overlap. Useful for RAG indexing and context previews.

Chunk size (chars)

Overlap (chars)

Document text

1 chunk(s)

Chunk #1 (94 chars)

# Doc

Paragraph one with enough text for RAG chunking.

Paragraph two continues the document.

Notes

RAG note

Min chunk size 100; overlap must be less than chunk size. Tune with semantic boundaries and embedding limits in production.

Set chunk size and overlap to split long Markdown or plain text. Preview RAG chunking before embedding and retrieval tuning.

Quick start

Set params

Chunk size ≥100 chars; overlap must be less than chunk size.
Paste document

Live preview of each chunk length and text.

Production tips

Character splits are a start; add semantic boundaries (headings, paragraphs) in production.

Features and use cases

Split Markdown by headings or max size while preserving structure for embedding or batch translation.

Use for technical doc RAG, long README chunking, and localization pipeline units.

Typical Workflow

After pasting long documents into the input box, set a chunk size of 500-1000 characters (ideal for most RAG systems). Click 'Split' to instantly preview chunks, with 10-20% overlap recommended for context continuity. Results allow copying single chunks or exporting all, ready for Embedding APIs or vector databases.

For code documentation, preserve Markdown syntax marks (like code block ``` symbols). Verify no accidental code truncation occurs, adjustable via chunk size or manual separators. Code docs may require smaller chunks (e.g., 300 chars) than plain text for logical integrity.

Examples

Example

Input

200 chars, 40 overlap

Output

Multiple indexed chunks

FAQ

Split by tokens?

Character-based; use Token Estimator for token budgets.

Why do my chunks contain truncated sentences?

This balances strict character limits. For complete sentences, increase chunk size or enable 'Prioritize Paragraphs' (slightly exceeds limits). Technical docs may tolerate minor truncations as RAG retrieval often recovers context via overlapping chunks.