AIBOX

Free Encoding and Data Tools: Base64, JSON and Hashes

Getting a value into exactly the shape a machine expects, and back again: Base64, URL encoding, hashes, JSON, CSV and identifiers that have to survive a round trip.

Encoding and data

Tools for the values you paste all day: the payload that will not parse, the string that has to survive a round trip unchanged, and the formats you have to hand back to a machine in exactly the shape it expects.

Encoding and decoding

Base64 encoder and decoder

Encode text to Base64 and decode it back, going through UTF-8 bytes so Chinese, Japanese and emoji work instead of throwing or quietly producing the wrong string. Accepts missing padding, line breaks and the URL-safe alphabet, and says so when valid Base64 decodes to bytes that are not text.

Runs in your browser · unlimited use

Data and formats

JSON formatter and validator

Paste JSON to pretty-print it with 2 spaces, 4 spaces or tabs, minify it, or find out why it will not parse — with the line and column of the error. Reports the traps that change your data, such as duplicate keys and integers too big to hold exactly.

Runs in your browser · unlimited use

What makes a browser the right place for this

Pasting a payload into a website usually means sending it somewhere. For a JSON formatter or a Base64 decoder that is a strange trade: both are pure operations on values you already have, they need nothing from a server, and the text is often the most sensitive thing on your screen — an API response with a token in it, a config file with a connection string, a log line with a customer identifier.

Everything in this category runs in the page you already loaded. There is no upload, no request to our server and no logging, because there is nothing to send: each page imports one small module of pure functions and works from there. It also means no usage limit, since a calculation that stays inside your browser costs us nothing to run — and it means the pages keep working when the network, or the VPN, does not.

What these tools refuse to guess

The most common feature in this category is the one we refused to build: silently deciding what an ambiguous value meant. A formatter that deletes a trailing comma or converts single quotes produces a file that looks correct in your editor and still fails in the program you send it to. An encoder that picks an alphabet on your behalf produces a string that decodes fine here and breaks in the query string it was meant for. Both hide the only fact worth knowing.

So the tools state their assumptions instead. The JSON formatter and validator reports the line, the column and a plain-language reason for a syntax error, and calls out the two cases that change your data even when the JSON is valid: duplicate keys, where every common parser keeps the last one, and integers beyond 253 − 1, where a long numeric ID is already rounded by the time anything can read it. The Base64 encoder and decoder goes through UTF-8 bytes so that Chinese, Japanese and emoji survive the round trip, accepts the unpadded, line-wrapped and URL-safe forms, and says so when valid Base64 decodes to bytes that are not text at all.

What belongs in this category next

The plan for this category is short and deliberately unglamorous: the conversions that come up while debugging, each of which can be done with browser-native APIs and no data of ours at all.

  • URL encoding — the other half of the same problem: a value that has to survive a query string, and the characters that do not survive it by default.
  • Hashes — SHA-256 and friends through the Web Crypto API, for checking that a file arrived intact.
  • CSV to JSON — the format most data actually arrives in, and the quoting rules that break naive parsers.
  • UUID and JWT — generating an identifier, and reading the claims out of a token without trusting it.

Each one ships in English and Chinese before it is listed here, which is why this page lists two tools rather than six. A link to something that does not exist yet is not a roadmap, it is a broken promise.

Which one do you need?

  • A string that looks like random letters, and you need to know what is inside it? Use the Base64 encoder and decoder: it goes through UTF-8 bytes, accepts unpadded, line-wrapped and URL-safe input, and tells you when the bytes are valid Base64 but not text.
  • A payload that will not parse, and you need to know where? Use the JSON formatter and validator: it gives the line, the column and a plain-language reason, and also reports duplicate keys and integers too large to hold exactly.
  • JSON that parses but is unreadable on one line? Same tool — choose 2 spaces, 4 spaces or a tab and press Format. Minify does the reverse when you need to paste it back into a request body.
  • A number from a log that is supposed to be a date? That one lives in the date and time category: use the Unix timestamp converter, which prints the unit it assumed and the other reading.
  • Text that needs counting rather than decoding? Use the word and character counter in the text category, which handles Chinese and Japanese by character instead of by whitespace.

Other tools and guides on this site

This is the category for values in transit; the rest of the site answers different questions.

  • All tools — every calculator and utility on the site, grouped by what it is for.
  • Date and time — timestamps, date gaps and working days, which is where the numbers in your logs usually point.
  • Text — counting and cleaning up text, for when the problem is the content rather than the encoding.
  • Editorial policy — how these pages are written and what we refuse to state without a citable source.
  • Privacy policy — what is and is not collected, and why the tools on this page never need to send anything.

Common questions

Is my JSON or my Base64 uploaded to your server?
No. Each page imports one small module of pure functions and does everything in the page you already loaded: parsing, formatting, encoding, decoding. There is no request to our server, nothing is logged, and closing the tab discards what you pasted. That also means no usage limit, since a calculation that never leaves your browser costs us nothing to run.
Why is this category called encoding and data instead of developer tools?
Because "developer tools" describes who you are rather than what the tools do, and it left no home for the ones that are not about code at all — Base64 inside an email header, JSON exported from an app, a CSV someone sent you. Encoding and data says what is actually here: getting a value into the exact shape something else expects, and back again.
Does Base64 hide or protect anything?
No, and treating it as protection is a real mistake. Base64 is a reversible encoding that makes arbitrary bytes survive a channel that only carries text; anyone who intercepts the string can decode it in one step. The page writes this out and names the padding, line breaks and URL-safe alphabet it accepted, so you can tell encoding apart from encryption.
Will the JSON formatter fix my broken JSON for me?
No, deliberately. A formatter that quietly deletes a trailing comma or converts single quotes makes a file look correct here and still fail wherever you send it next, while hiding what the problem really was. The tool reports the line, the column and a plain-language reason, and leaves the decision about what the data was meant to be to you.
What belongs in this category next?
URL encoding and decoding, hashes such as SHA-256 through the browser's own Web Crypto API, CSV converted to JSON and back, UUID generation and JWT decoding. All of them are browser-native operations with no data of ours behind them, and each ships in both languages before it appears here rather than being promised as a link.

Every calculator on this page runs in your browser: nothing you type is sent to us. See our privacy policy.