JWT Decoder

The Complete JWT Guide

The three-part JWT structure, what the most common claims mean, and the critical difference between decoding a token and verifying it.

Open JWT Decoder

The Three-Part JWT Structure

A JSON Web Token is a single string made of three parts separated by dots (.): header.payload.signature. Each part is encoded separately using base64url (a URL-safe variant of base64 that replaces + with - and / with _, typically without padding) before being joined together.

PartContainsPurpose
HeaderA small JSON object, e.g. {"alg":"HS256","typ":"JWT"}Describes the token type and the signing algorithm used
PayloadA JSON object of claims - who, when, whatCarries the actual data the token is asserting
SignatureA cryptographic signature over the header and payloadLets a server with the right key confirm the token wasn't altered

Because the header and payload are only base64url-encoded, not encrypted, they're plain text to anyone who decodes them - that's the whole point of this tool, and exactly why a JWT should never carry secret data in its payload.

Common JWT Claims Explained

sub (Subject)Identifies the principal the token is about - typically a user ID or username the token was issued for.
iat (Issued At)A Unix timestamp (seconds since 1970) marking exactly when the token was created by the issuer.
exp (Expiration)A Unix timestamp after which the token must be rejected, even if the signature is still valid - limits how long a stolen token stays useful.
nbf (Not Before)A Unix timestamp before which the token must NOT be accepted yet, even if everything else checks out - used to issue a token that only becomes valid at a future time.
aud (Audience)Identifies the intended recipient(s) of the token, such as an API or application - services should reject tokens meant for someone else.
iss (Issuer)Identifies who created and signed the token - a server should only trust tokens from issuers it explicitly recognizes.
jti (JWT ID)A unique identifier for the token itself, sometimes used to detect replay or to support revocation lists.

Decoding vs. Verifying: The Critical Distinction

It's easy to conflate "I can read this token" with "this token is genuine" - but they are two completely different operations. Decoding is simply reversing the base64url encoding to reveal the JSON that was already sitting inside the token in plain sight; anyone, including an attacker, can do this with no special knowledge at all. This tool - and every purely client-side JWT tool - only decodes.

Verifying is cryptographically proving the signature was produced by the real issuer and that the header/payload haven't been tampered with since. That requires the secret key (for symmetric algorithms like HS256) or the issuer's public key (for asymmetric algorithms like RS256), neither of which lives anywhere near your browser - it stays on the issuing server, exactly where it's supposed to. So a decoded payload is safe to inspect and debug, but it must never be treated as trusted or authentic input for a security decision until a server with the signing key has verified it.

Common JWT Use Cases

Session TokensAfter login, a server issues a signed JWT the browser stores and sends back on subsequent requests to prove who's logged in without re-checking a password every time.
API AuthorizationA bearer token attached to the Authorization header lets an API verify the caller's identity and permissions on every request, statelessly.
Single Sign-On (SSO)Identity providers issue JWTs (often ID tokens in OpenID Connect) that let one login work across multiple connected applications.
Short-Lived Action LinksPassword reset or email-confirmation links can embed a signed, expiring JWT instead of a database-backed one-time code.

Frequently Asked Questions

Does decoding a JWT prove it's genuine?

No. Decoding only reveals the JSON already sitting inside the token - anyone can do it, including an attacker. Proving a token is genuine requires verifying its signature with the issuer's secret or public key, which never reaches this browser-only tool.

Does this tool send my JWT to a server?

No. Everything is decoded entirely in your browser using JavaScript - your token, including any sensitive claims inside it, never leaves your device.

What do the sub, iat, and exp claims mean?

sub identifies who the token is about, typically a user ID. iat records when the token was issued, as a Unix timestamp. exp is the Unix timestamp after which the token must be treated as expired and rejected, even if its signature is still valid.

What's the difference between aud and iss?

iss (issuer) identifies who created and signed the token. aud (audience) identifies who the token is intended for. A service should check both - reject tokens from an untrusted issuer, and reject tokens meant for a different audience than itself.

Why can I read a JWT's payload without a password?

Because the header and payload are only base64url-encoded, not encrypted - anyone can decode them, which is exactly what this tool does. Only the signature requires a secret key, and that's what makes the token trustworthy, not the payload itself. JWTs should never contain secret data in the payload.

What are common uses for JWTs?

JWTs are most often used as session/authentication tokens after a user logs in, and as bearer tokens for authorizing requests to an API - the server issues a signed token once, and every subsequent request includes it instead of re-sending credentials.

Open JWT Decoder