Paste Web Content
What to Clean
Everything is on by default — toggle off anything you'd rather keep.
Clean Text
Strip HTML, Smart Quotes & Invisible Characters • See the Token Savings
Everything is on by default — toggle off anything you'd rather keep.
Text copied from a web page rarely comes through clean — it often drags along HTML tags, encoded entities like , curly "smart" quotes, em dashes, non-breaking spaces, and invisible zero-width characters that add length without adding meaning. All of that costs tokens when you paste it into an AI tool, without helping the model understand your content any better. This tool strips HTML using the browser's own parser (far more reliable than regex for handling real-world, sometimes malformed HTML), normalizes typographic punctuation to plain ASCII, removes invisible characters, and collapses excess whitespace — then shows you the actual character and token savings using a real BPE tokenizer, the same algorithm behind OpenAI's tiktoken. Everything runs locally in your browser; nothing you paste is ever uploaded anywhere.