BaseKit - free, no upload, runs in your browser

Base64 encode and decode, with full UTF-8 support

BaseKit encodes and decodes Base64 with correct UTF-8 handling - so emoji, Chinese text, and accented characters survive the round trip - plus a URL-safe mode that strips the + / = characters for tokens and query parameters. Everything runs in your browser.

• Full UTF-8 support: Chinese, emoji, and other multibyte text encode and decode correctly
• URL-safe output mode for JWT segments and query parameters (no +, /, or =)
• One-click swap moves text and encoded output between panes
• Encode and decode happen locally - sensitive payloads and tokens never leave your machine

The UTF-8 trap most Base64 tools get wrong

Many online Base64 tools convert plain ASCII and silently corrupt anything else. Encode 中文 or an emoji and the round trip comes back as mojibake. BaseKit encodes and decodes with a UTF-8-safe codec, so multibyte text survives intact - the same behavior you get from standard libraries in Node, Python, and Go.

That makes it safe for real payloads: JWT header and payload segments, SAML assertions, small image data URLs, and API keys copied out of logs.

When to use URL-safe Base64

Standard Base64 uses + and /, which are not URL-safe, and = padding that complicates string comparison. URL-safe Base64 swaps in - and _ and drops the padding, which is why JWT segments and many API tokens use it. BaseKit produces both from the same input, so you can compare a token in a query string against a decoded value without transcription errors.

How to encode or decode

• Open BaseKit and paste plain text on the left or Base64 on the right.
• Press Encode or Decode - toggle URL-safe when working with tokens or query parameters.
• Use Copy result to grab the output, or Swap to flip the panes for the reverse direction.

Frequently asked questions

Does it handle Chinese text correctly?

Yes. BaseKit uses a UTF-8-safe codec, so 中文, emoji, and accented characters round-trip exactly instead of becoming mojibake.

What is URL-safe Base64?

A variant that replaces + and / with - and _, and drops = padding. It is used by JWTs and API tokens so the string is safe inside URLs and query strings.

Can it encode or decode files?

Text is supported directly. For files, base64 encoding is a separate concern - paste the base64 text and use the decode side to inspect it.

Is my input uploaded?

No. Encoding and decoding run entirely in your browser, which matters because people often base64-decode JWT segments and tokens.

What is the size limit?

Up to 2 MB of text per operation, which is far beyond typical token and payload sizes.

Related local tools

• IDMintGenerate UUIDs in bulk, locally
• RegexLabTest regular expressions with live highlighting
• RedactLabScrub emails, tokens, and IPs from text