What does a ULID decoder actually pull out of the ID?

Two things, plus the readable times. A ULID is 26 Crockford Base32 characters; the first 10 are a 48-bit millisecond timestamp and the last 16 are 80 bits of randomness. Paste 01ARYZ6S41TSV4RRFFQ69G5FAV and the decoder returns the timestamp as 1469918176385 ms, renders it as 2016-07-30T22:36:16.385Z, and shows the random tail TSV4RRFFQ69G5FAV on its own. The split is fixed at 10 + 16, so the decode is unambiguous.

How is a ULID structured, time versus randomness?

26 characters total: 10 of timestamp followed by 16 of randomness. Crockford Base32 stores 5 bits per character, so 10 chars hold 50 bits, of which the high 48 are the millisecond time (the first character uses only 3 of its 5 bits, which is why a valid ULID never starts above 7), and 16 chars hold the 80 random bits. 48 + 80 = 128 bits — the same payload as a UUID, just encoded more compactly and without dashes.

Why is the alphabet Crockford Base32 and not ordinary Base32?

Crockford Base32 drops the characters that humans confuse: I, L, O and U are not in the alphabet, so there is no mixing up I with 1 or O with 0, and U is removed to avoid accidentally spelling words. The 32 symbols are 0-9 then A-Z minus those four. A correct decoder rejects any ULID containing I, L, O or U; if you paste one with those letters this tool flags it as invalid rather than silently mis-decoding.

How do I get the creation time out of a ULID?

Take the first 10 characters and read them as a big-endian Crockford Base32 number; that value is the milliseconds since the Unix epoch. For 01ARYZ6S41 the math gives 1469918176385 ms, which is 2016-07-30 22:36:16.385 UTC. You do not need to query a database or a created_at column — the time is baked into the ID, and this page does the base-32 arithmetic and the date formatting for you, in UTC, ISO and your local zone.

What is the difference between a ULID and a UUID for decoding?

A UUIDv4 is fully random, so there is nothing meaningful to decode out of it — no time, no order. A ULID embeds its creation timestamp in the leading characters, so it is both decodable and lexicographically sortable: sort ULIDs as plain strings and you sort them by time. Both are 128-bit IDs, but a ULID is 26 characters with no dashes versus a UUID's 36, and only the ULID lets you recover when it was minted.

Is decoding a ULID safe, and does it leak anything?

Decoding here is fully local, so the ULID you paste is never uploaded or logged. Be aware of what the decode reveals, though: the first 10 characters expose the exact millisecond the ID was created, so a ULID shown in a public URL leaks roughly when that record was made, and a run of adjacent ULIDs can hint at how fast records are being created. The 80 random bits stay unguessable, so the ID is not predictable, but treat the embedded time as public if the ID is.

Find when a record was created from its ID alone

A support ticket quotes a row ID like 01ARZ3NDEK... and you need to know when it was created without database access. Paste it into the decoder and read the exact millisecond plus a readable UTC and local time. The timestamp is baked into the first 10 characters, so there is no created_at lookup and no query — the ID answers the question by itself.

Triage a log file full of ULIDs at once

You have a log dump where every line is a ULID and you want the timeline. Paste the whole block, one ULID per line, and the batch view decodes every row to its timestamp and UTC time. Bad or truncated lines are flagged as invalid instead of silently mis-decoding, so you can spot the corrupt entries while you read off the order.

Verify a ULID is well-formed before storing it

An upstream service hands you IDs and you suspect some are malformed. Paste one in: the decoder rejects any string that is not 26 Crockford Base32 characters, that contains I, L, O or U, or whose first character is above 7 (which would overflow the 48-bit time). A clean decode means it is a valid ULID; an error means reject it at the door.

Confirm two ULIDs are in the order you expect

You rely on ULIDs sorting by time but a result set looks off. Decode the two IDs in question and compare their timestamps directly — the milliseconds tell you which was minted first, independent of any string comparison. It is the fast way to confirm that ORDER BY id really is giving you creation order before you debug further up the stack.

ULID Decoder — Extract Timestamp + Randomness from a ULID

Pull the timestamp and randomness back out of a ULID — decode one or paste many, with UTC, ISO and local time — 100% in your browser

Runs locally
Category Developer & DevOps
Best for Formatting, validating, shrinking, or inspecting code-adjacent text.

A ULID is a 26-character Crockford Base32 string: the first 10 characters are a 48-bit millisecond timestamp, the last 16 are 80 bits of randomness. Because the time sits in the leading characters, ULIDs sort by creation time with a plain string sort. Paste one ULID — or many, one per line — and read its timestamp and random tail. Everything decodes in your browser.

ULID(s)

Multiple lines decode at once — one ULID per line.

Decoded

Paste a ULID above to split out its timestamp and randomness.

Tip: a plain alphabetical sort of ULIDs is also a sort by creation time, since the timestamp lives in the leading characters.

What this tool does

Decode a ULID in your browser: paste any 26-character Crockford Base32 ULID and read its embedded creation time and random tail. A ULID splits, most-significant first, into a 48-bit millisecond timestamp in the first 10 characters and 80 bits of randomness in the last 16, so this tool is the exact inverse of generation: it pulls the timestamp out as a Unix-ms number and renders it as ISO 8601, a readable UTC time and your local time, then shows the 16-character random part on its own. It validates the Crockford alphabet (no I, L, O or U), rejects a wrong length, and refuses a first character above 7 that would overflow the 48-bit time, so a mistyped ID is caught immediately. Paste many ULIDs, one per line, to decode a whole batch from a log file at once, and copy any value with one click. Because the timestamp lives in the leading characters, a plain alphabetical sort of ULIDs is also a sort by creation time. Nothing is uploaded; all decoding runs locally.

Tool details

Input: Text; The page exposes text boxes, numeric controls, file pickers, or structured inputs depending on the tool.
Output: Live result + Copy; The result area focuses on usable output, with copy, download, or preview actions when supported.
Privacy: Browser-side processing; The main tool logic does not call an external API, so inputs normally stay in the current tab.
Save / share: Shareable URL state; Key settings are encoded in the URL so another person can reopen the same setup.
Performance budget: Initial JS <= 9 KB; No WASM budget is declared, keeping the tool quick to open on mobile.
Best fit: Developer & DevOps · Developer; Category and role tags drive related tools, internal links, and quick fit checks.

How to use

1. Input

Paste or drop your content into the tool panel.
2. Process

Click the button. All processing is local in your browser.
3. Copy / Download

Copy the result or download to disk in one click.

How ULID Decoder fits into your work

Use it in the small gaps between coding, reviewing, debugging, and shipping.

Developer jobs

Formatting, validating, shrinking, or inspecting code-adjacent text.
Preparing snippets for documentation, tickets, commits, or handoff.
Checking a small payload quickly without switching tools.

Developer checks

Run irreversible transforms like minify or obfuscate on a copy.
Keep secrets out of pasted snippets unless the tool explicitly stays local.
Use your normal tests or linter before shipping transformed code.

Good next steps

These links move the current task into a more complete workflow.

Real-world use cases

Find when a record was created from its ID alone
A support ticket quotes a row ID like 01ARZ3NDEK... and you need to know when it was created without database access. Paste it into the decoder and read the exact millisecond plus a readable UTC and local time. The timestamp is baked into the first 10 characters, so there is no created_at lookup and no query — the ID answers the question by itself.
Triage a log file full of ULIDs at once
You have a log dump where every line is a ULID and you want the timeline. Paste the whole block, one ULID per line, and the batch view decodes every row to its timestamp and UTC time. Bad or truncated lines are flagged as invalid instead of silently mis-decoding, so you can spot the corrupt entries while you read off the order.
Verify a ULID is well-formed before storing it
An upstream service hands you IDs and you suspect some are malformed. Paste one in: the decoder rejects any string that is not 26 Crockford Base32 characters, that contains I, L, O or U, or whose first character is above 7 (which would overflow the 48-bit time). A clean decode means it is a valid ULID; an error means reject it at the door.
Confirm two ULIDs are in the order you expect
You rely on ULIDs sorting by time but a result set looks off. Decode the two IDs in question and compare their timestamps directly — the milliseconds tell you which was minted first, independent of any string comparison. It is the fast way to confirm that ORDER BY id really is giving you creation order before you debug further up the stack.

Common pitfalls

Validating a ULID with the full A-Z alphabet. Crockford Base32 excludes I, L, O and U, so a regex that allows every letter will accept malformed IDs and silently mis-decode them. Check against the exact 32-symbol set, which is what this tool does — paste one with an I or O and it flags the error instead of guessing.
Reading the timestamp out of the wrong end. The time is the first 10 characters, not the last; the trailing 16 are pure randomness with no time in them. Slicing from the back, or treating the whole 26 characters as the time, gives a nonsense date. The split is always 10 timestamp + 16 randomness.
Treating the random tail as a secret or a checksum. The 80 random bits make collisions effectively impossible, but they are not a signature and they verify nothing. A ULID is an identifier, not a token — never rely on its randomness as a security check, and remember the timestamp half is fully readable by anyone who has the ID.

Privacy

Every decode runs in your browser tab — the Crockford Base32 arithmetic, the timestamp extraction and the date formatting are all plain JavaScript with no network call. A ULID you paste to inspect is never uploaded or logged, and batch input stays entirely local. One thing to weigh: the shareable URL encodes whatever you typed in the query string, so a "share link" pasted into chat will record those ULIDs in the recipient server's access log. Since a ULID reveals its creation time, use the copy button instead of the URL when that timing is sensitive.

FAQ

Tool combos

Folks in your role tend to reach for these alongside this tool.

Developer

Browse all tools for this role

ULID Decoder — Extract Timestamp + Randomness from a ULID

What this tool does

Tool details

How to use

1. Input

2. Process

3. Copy / Download

How ULID Decoder fits into your work

Developer jobs

Developer checks

Good next steps

Real-world use cases

Find when a record was created from its ID alone

Triage a log file full of ULIDs at once

Verify a ULID is well-formed before storing it

Confirm two ULIDs are in the order you expect

Common pitfalls

Privacy

FAQ

JWT Decoder

File Hash Calculator

ULID Generator

UUID Generator

Snowflake ID Decoder

Unix Timestamp Converter

AI Eval Planner

API Key Generator

ASCII Table Generator

ASCII Table Reference

awk + sed Cheatsheet

AWS CLI Cheatsheet