What is Base64 Decoding?
Base64 decoding reverses the encoding process: a string of Base64 characters is converted back into the original byte sequence, then interpreted as text. Developers encounter encoded strings in JWT payloads, Basic auth headers, database fields, and log files.
Vertex Solutions' decoder assumes the underlying bytes represent UTF-8 text — the same assumption used during encoding in the companion Base64 Encode tool.
Reading Encoded Data
When debugging an API, you might receive a field like "data": "SGVsbG8gV29ybGQ=". Pasting that value here reveals Hello World. For nested encodings (Base64 inside JSON inside Base64), decode one layer at a time and inspect the result before proceeding.
Log analysis often surfaces Base64 blobs that are actually gzip-compressed JSON. If decoded text still looks like garbage, the payload may need decompression or a different character encoding — not further Base64 decoding.
URL-Safe and Padding Variants
RFC 4648 defines standard Base64 (+, /) and URL-safe Base64 (-, _). JWTs and many web APIs use the URL-safe variant without padding. Before decoding:
- Replace
-with+and_with/. - Append
=characters until the string length is divisible by 4.
Standard decoders, including browser atob(), require valid padding for some inputs.
Recovering Original Text
After successful decoding, copy the plain text for use in editors, diff tools, or further processing. If you need to re-encode after editing, switch to Base64 Encode to produce a fresh encoded string.
Remember: decoding exposes whatever was encoded. If someone Base64-encoded a password (a common but insecure practice), the plaintext is immediately visible after decoding.
Worked Examples
Example 1 — Standard padded Base64 string
- Input:
SGVsbG8gV29ybGQ= - Decoded output:
Hello World
Example 2 — Inspecting a JSON payload (such as a JWT middle segment)
- Input:
eyJ1c2VySWQiOjEwNCwicm9sZSI6ImVkaXRvciJ9 - Decoded output:
{"userId":104,"role":"editor"}(paste the result into the JSON Formatter or JSON Validator to pretty-print and check syntax).
Example 3 — UTF-8 accented characters
- Input:
Q2Fmw6k= - Decoded output:
Café
If you need to re-encode edited text back into Base64, switch to Base64 Encode.
When Decoding Fails
| Symptom | Likely cause |
|---------|----------------|
| Invalid character error | URL-safe variant (-, _) or stray whitespace in the string |
| Garbled Unicode output | Wrong byte interpretation — input may be binary, not UTF-8 text |
| Empty output | Input was only padding characters or an empty string |