Logo

Decrypt Text Online

Paste the string. We check what is actually wrong before asking you to guess at the password again.

Everything above happens in this tab. The string is never uploaded, which also means we cannot help you if the password is gone — there is no copy of it anywhere. When the password is simply gone covers what is and is not worth trying next.

A string arrives, you paste it in, and a red line says decryption failed. You try the password again, more carefully, and it fails again. Most of the time the password was never the problem — so this page checks the data first and tells you what it found.

Why won't my encrypted text decrypt?

Four things go wrong, and only one of them is the password. Knowing which saves you from typing the same eleven characters twenty times.

The string may be truncated, because something between the sender and you had a length limit. Its characters may have been altered, because a mail client wrapped the lines or a phone keyboard capitalised something. It may have come from a different tool entirely. Or the data may be perfect and the password wrong.

The first three are structural, which means we can prove them before you type anything. Paste a string above and the verdict appears straight away, with no password and no waiting. When the data is sound we say so plainly, and then the password really is the only variable left.

How to decrypt a message someone sent you

Get the string in one piece first. That sounds obvious and it is the single biggest cause of failure we see.

Select from the very first character. Our output begins with OWK1, and a paste that starts partway through looks like valid base64 with no header — which we detect and call out, because it is such a common accident. If the sender used a chat app, ask them to send it as a file instead.

Then get the password from a different channel than the message. If both arrived in the same thread, the encryption did very little for either of you, and getting the key to the other end is worth reading before the next one.

What AES decrypt with a password is doing while you wait

The pause before the text appears is not the decryption. Decrypting a paragraph is instant; the wait is your password being turned into a key.

That step runs a hashing function hundreds of thousands of times on purpose. It costs you a fraction of a second once and it costs an attacker the same fraction on every single guess, which is what turns a short password from hopeless into merely risky.

The iteration count is stored inside the string rather than assumed, so files made with our older settings still open with those settings. It also means the pause is longer on newer files than old ones, which is the correct direction for it to move.

The failure we refuse to guess at

When the data is intact and the password is wrong, we say the password is wrong or the data was altered — and that we cannot tell which. That is not hedging, it is the actual limit.

AES-GCM checks the whole message against an authentication tag before releasing any of it. A wrong key produces a failed tag. A single flipped bit produces a failed tag. The two are indistinguishable by design, and a mode that let you tell them apart would leak information to someone probing your data. Why GCM beats CBC goes through what that tag buys you and what it costs.

In practice it is the password nearly every time. We still will not print that as a diagnosis, because the one occasion it is wrong is the occasion someone needed to know their file had been tampered with.

Opening files from the older vault format

Files that begin with STVAULT1.0 came from an earlier version of our own tool. They still open here, and we plan to keep it that way.

Two of them get a warning rather than a clean bill. If a file opens with no password at all, it was saved in a mode that used a fixed key published in the page's own code — it was never protected from anybody, and we say so instead of quietly handing over the text.

If the file carries an old recovery block, it holds the password encrypted under three security answers. That made the whole thing only as strong as a first pet's name. Decrypt it, re-encrypt with the encrypt text online page, and delete the original.

When there is nothing left to try

Sometimes the answer is that the text is gone. We would rather tell you in a minute than let you spend a week on it.

There is no master key, no copy on a server, and no support route that ends in recovery. The deliberate slowness of key derivation also means brute force is not realistic beyond a handful of remembered variations. When the password is simply gone sets out what is genuinely worth attempting and which services are selling hope.

Decrypt text online without guessing at the cause

Paste, read the verdict, then decide. If the string is broken you have learned something in two seconds; if it is intact you know the password is the only thing between you and the text.

For whole files rather than pasted strings, our encrypting a whole file page handles the same job with the original filename preserved. And if what you are protecting is a plain text document on your own machine, a plain .txt file on your own disk covers the operating-system routes and their sharp edges.

Frequently Asked Questions

Why can you not tell me whether it is the password or the file?

+
Because the check that fails is the same check in both cases. The mode we use verifies the whole message against a tag before releasing any of it, and a wrong key produces exactly the same failure as an altered byte. Anyone claiming to distinguish the two is guessing, so we say which of the other problems we have ruled out instead.

The tool says part of my text is missing. What usually causes that?

+
A length limit somewhere between the sender and you. Chat apps truncate very long strings, some ticketing systems cut at a field width, and pasting from a terminal can drop everything past the visible buffer. Ask for it again as a file attachment rather than as a message.

It says characters were changed in transit, but I only copied and pasted.

+
Copying is where it happens. Mail clients wrap long lines, note apps convert straight quotes to curly ones, and phone keyboards capitalise the first letter of what looks like a sentence. Any of those alters the data, which is why sending it as an attachment is more reliable than pasting it.

I have an old file from your vault tool. Will it still open?

+
Yes, and we intend to keep it that way. We still read the older STVAULT1.0 format, decrypt it with the password you used, and then tell you to re-encrypt. Dropping support would have destroyed real data to tidy up our own mistake.

Why did an old file open without asking me for a password at all?

+
Because it was never protected. The older tool let you encrypt without a password and quietly used a fixed key that shipped in the public code, while still calling the result encrypted. If a file opens that way here, assume anyone who had a copy could read it.

My old file had security questions attached. Is that a problem?

+
It is, and we flag it when we see it. That block holds your password encrypted under three answers, so the file was only ever as strong as those answers. Decrypt it, re-encrypt it with a real password, and delete the original.

Can I brute-force a password I half remember?

+
Not here, and not quickly anywhere. Each attempt has to run the full key-stretching step, which takes a noticeable fraction of a second by design, so a script gets through very few guesses per second. If you remember the structure and only a couple of characters are uncertain, trying by hand is realistic. Beyond that it is not.

Does decrypting something here leave it anywhere?

+
No. It lives in the page while the tab is open and goes when you close it, and nothing is written to storage or sent anywhere. Bear in mind that your clipboard is a different matter — copying the plaintext puts it somewhere other apps can read.

Someone sent me a string that is not from your tool. Can you open it?

+
No, and we will say so immediately rather than let you try passwords. Encrypted output is not a universal format: the cipher, the mode, the key-stretching settings and the layout all differ between tools. Whoever encrypted it needs to tell you what they used.

The text decrypted but looks like nonsense. What happened?

+
That almost never happens with this format, because a wrong password fails the tag check rather than producing garbage. If you do see scrambled output, the likely explanation is that the original text was itself encoded or compressed before being encrypted. Check with whoever sent it.

Related Tools You Might Like

Advertisement