JWT Decoder
Decode a JWT payload locally. This does not verify its signature.
Enter a JWT
Decode a JWT payload locally. This does not verify its signature.
Enter a JWT
A JSON Web Token is three Base64URL sections separated by dots: a header, a payload, and a signature. The payload holds the claims, such as who the token is about, who issued it, and when it expires. This tool reads the middle section, decodes it, and prints the claims as indented JSON so you can see what a token actually says.
Decoding is not verification. This page does not check the signature, does not confirm who issued the token, and does not test whether it has expired. Anyone can create a token with any claims they like, so a readable payload proves nothing about trust. Signature checks must be done by your server or a proper library using the correct key.
The decoder only shows the payload. The header, which names the signing algorithm, is not displayed, and the signature is ignored. That keeps the output focused on the part you usually need when debugging login problems, missing roles, or wrong audience values.
A user is redirected back to a sign-in page. Decoding the access token can reveal an expired exp value, a wrong aud, or a missing scope, so you know whether the fault lies in the token or somewhere else.
After changing roles, groups, or custom claims in your identity provider, decode a fresh token to confirm the new values appear in the payload before you spend time debugging the application.
For a course or a team onboarding, paste a sample token and show that the contents are only encoded, not encrypted. It makes clear why secrets should never be stored in a JWT payload.
A support ticket includes a token copied from a request log. Decoding it shows the subject, issue time, and expiry without writing a throwaway script, which helps you line it up with the incident timeline.
Decoding is done in your browser with built-in functions. The token you paste is not sent to a server by this tool, though you should still treat real tokens as sensitive.
No. It only decodes the payload so you can read it. Verification needs the signing key or public key and must run in your own backend or a trusted library. A payload that decodes cleanly can still be forged, so never use this output to decide whether a token is genuine.
A standard signed JWT is not encrypted. Its header and payload are only Base64URL encoded, which anyone can reverse, as this page shows. The signature protects against tampering, not against reading. Do not put passwords or private personal data in claims.
Look for the exp claim, a Unix time in seconds. If it is earlier than the current time, the token has expired. This tool shows the number only, so convert it to a date yourself or use the Unix Timestamp Converter on a date you compute.
These are registered claims. iss is the issuer, sub is the subject, usually a user ID, aud is the intended audience, exp is the expiry time, and iat is when the token was issued. Applications can add custom claims such as roles or email.
The token may be cut off, may contain extra characters or quotes, or may not be a three-part JWT. The payload must also be valid JSON after decoding. Opaque tokens and encrypted tokens do not follow this structure and cannot be read here.
The decoding happens in your browser and the token is not sent to a server by this tool. Even so, a live token grants access until it expires, so prefer test tokens, avoid sharing screenshots, and revoke any token you think was exposed.