Find hidden messages in text

Some text carries a second, invisible message that only software can read. Paste it here to decode the message and strip it out.

Checked in your browser. Nothing is uploaded. Turn off your Wi-Fi and it still works.

Paste your text

Under the light
Inspection report—
    Cleaned text

    Ticked items are cleaned. This changes how the text is stored, not who wrote it.

    How it works

    Short answer

    Text can hide a message using Unicode tag characters (U+E0000 to U+E007F), which are invisible but each stand for a normal letter. Paste the text above and the decoder shows the hidden message in plain words, then gives you a clean copy without it.

    Text: characters that take up no space

    ON SCREEN results IN THE FILE r e s u l t s U+200B · 3 bytes · 0 px wide

    A zero-width space sits between letters but draws nothing. It can break search, word counts and code, and can be used to mark copied text. Tag characters go further and spell out invisible messages.

    Three ways text hides a message

    1. Tag characters. Each visible letter has an invisible twin in the range U+E0020 to U+E007E. A string of them can spell out anything. This is the easiest kind to decode, and it's what the example above uses.
    2. Variation selector runs. Variation selectors normally pick the style of an emoji. Long runs of them can carry one byte each, so a single emoji can secretly hold a sentence.
    3. Zero-width binary. Patterns of zero-width spaces and joiners can stand for ones and zeros. These are flagged as invisible characters.

    Why this matters for AI

    Hidden text is a known way to attack AI assistants. Someone hides an instruction like ignore previous instructions inside a web page or email. A person sees nothing odd, but a chatbot reading the text sees the instruction. If you're pasting text from an unknown source into an AI tool, cleaning it first removes anything hidden.

    Questions

    How can a message be invisible?

    Unicode has a block of tag characters that mirror the normal letters but have no visible shape. A program can turn any message into these characters and slip them into a sentence. Copying the sentence copies the message too.

    Who uses hidden text messages?

    They're used to trace leaked documents, to watermark copied text, and in prompt-injection attacks that hide instructions for AI chatbots inside normal-looking text.

    Can it find messages hidden in zero-width spaces?

    It finds every zero-width character and flags long runs of them. Tag characters and variation-selector runs are decoded into readable text when the bytes form one.

    Is the message removed when I clean the text?

    Yes. With Hidden messages ticked, the cleaned copy has none of the invisible characters left.