Built for LLM workflows

Sentence Splitter for LLMs

Split your text into clean, model-ready sentences for RAG, training, and evaluation.

0 / 10,000 characters

Tip: Drag and drop multiple .txt, .md, or .csv files to populate the input.

Output

Run the splitter to see JSON and plain text output formats.

Chunk preview

Visualize approximate chunking with token limits and overlap.

Token counts are approximated by whitespace splitting for quick insight.

Run the splitter to explore how chunks are distributed.

How it works

  • Trim the input and optionally collapse line breaks.
  • Split sentences with regex on `.`, `!`, `?` followed by whitespace.
  • Remove trailing punctuation when toggled off.