Split Text by Token Limit
Cut long Markdown or prompts into chunks under a maximum token budget. Runs locally — no upload.
All processing on this page runs locally in your browser unless noted otherwise.
Splitting runs locally in your browser. Your text never leaves your device.
Source
Chunks (0)
No chunks yet.
How to use
- Paste a long document or load a local file.
- Choose a model and max tokens per chunk.
- Copy individual chunks or export a Markdown report.
FAQ
How does splitting work?
Text is encoded with the selected model’s tokenizer (OpenAI exact; others via local o200k proxy). Chunks are cut at the token limit, preferring paragraph/newline boundaries when they fit.
Will chunks always be under the limit for Claude / Gemini?
Only approximately. Estimate models use a proxy tokenizer — leave a safety margin (e.g. 10–15%) if you need headroom for the vendor’s real tokenizer.
Is my document uploaded?
No. Splitting runs entirely in your browser.