RAG Text Chunk Splitter
NewPreview how a document is split into chunks for retrieval-augmented generation, with adjustable chunk size, overlap and recursive or fixed-size strategies.
Frequently Asked Questions
What chunk size should I use for RAG?
A common starting point is 300–1000 characters (roughly 100–250 tokens) with 10–20% overlap. Test against real queries, since the best size depends on your content and embedding model.
What is chunk overlap for?
Overlap repeats the end of one chunk at the start of the next so sentences split at a boundary still keep their context.
Recursive vs fixed-size?
Recursive splitting tries paragraph, line, sentence and word boundaries in order, keeping text coherent. Fixed-size cuts at exact character counts.
Are token counts exact?
No, they are estimated at about 4 characters per token. Use your model's tokenizer for exact numbers.
Related Tools
LLM Token Counter
NewEstimate token counts for GPT, Claude, and Llama-style models as you type.
AI Cost Calculator
NewEstimate LLM API costs from input/output token counts across popular models.
Prompt Template Builder
NewBuild reusable AI prompt templates with {{variable}} placeholders and instant preview.