Encoder Workbench
Chain Base64, Base32, hex, URL, HTML entity, ROT13 and compression steps, and see each intermediate result as text or hex.
Sidebar group: Analysis · Works offline: yes (nothing is executed or rendered) · Needs admin rights: no
When to use it
- A log field, URL parameter or alert holds an encoded blob and you need to peel it back one step at a time.
- You want to check what a value looks like after URL encoding or Base64 before putting it into a config, API call or test.
- A payload is Base64 of gzip of something else, and you want to see each stage instead of guessing.
- The Script Decoder could not resolve a custom chain and you want to decode the payload by hand.
Quick start
- Open Encoder Workbench from the Analysis group.
- Paste the text into Input. Tick input is hex bytes if you are pasting hex.
- Pick a step from + add step..., for example Base64 decode.
- Read the step's result. If the bar under the last step says Looks like it could be:, click a suggestion to add that step.
- Keep adding steps until the result is readable. Reorder with the arrows or remove a step with the cross on its chip.
- Click Copy result for the final output, or Export for a report.
Inputs and options
| Field | What it means | Default / accepted values |
|---|---|---|
| Input | Text to encode or decode. | Up to 5 MB. |
| input is hex bytes | Treat the input as hex bytes instead of text. | Off. Spaces are allowed in the hex. |
| + add step... | Appends a step to the chain. | See the list below. Up to 30 steps. |
| Step chips | Each step in order, with move up, move down and remove buttons. | |
| Clear | Removes all steps. | Shown when there are steps. |
| Text / Hex (per result) | Switches a result between text and a hex dump. | Shown when the result is valid text. |
| Suggestions | One-click steps guessed from what the final result looks like. | |
| Copy result | Copies the final result (text, or hex for binary output). |
Available steps:
- Decode: Base64 decode, Base64url decode, Base32 decode, Hex decode, Base85 decode, URL decode, URL decode (+ as space), HTML entity decode.
- Encode: Base64 encode, Base64url encode (no padding), Base32 encode, Hex encode (lowercase), URL encode, HTML entity encode.
- Other: ROT13, Gzip decompress, Zlib/deflate decompress, XZ/LZMA decompress, Gzip compress.
The chain recalculates about 200 ms after any change. Base64 decoding is lenient: characters outside the alphabet are dropped and padding is optional.
Reading the results
With no steps, the input is shown with its byte count. With steps, each one gets a box headed by its number and name:
- The byte count of its output and a Text / Hex toggle.
- The output as text, or "Binary content, shown as hex." with a hex dump (offset, hex bytes, ASCII) when it is not valid text.
- A note when something was partly handled, for example "N unrecognized named reference(s) left as written" after HTML entity decode.
- A hint when the output looks like a JWT or a SAML message (see below).
- If a step fails, its box shows Stopped here: with the reason, and later steps are not run.
Below the last successful step, Looks like it could be: offers next steps based on the bytes: Gzip, XZ or Zlib decompress from their magic bytes, Hex decode for even-length hex, Base64 or Base64url decode for strings in those alphabets, URL decode when %XX escapes are present, and HTML entity decode when entities are present.
Example: input H4sIAAAAAAAA/... (Base64 text). Add Base64 decode: the result is binary and shown as hex starting 1f 8b, and the bar suggests + Gzip decompress. Add it: the result is readable text, for example GET /login?next=https%3A%2F%2Fapp.example.com%2F, and the bar now suggests + URL decode. One more click gives GET /login?next=https://app.example.com/.
What the results do and do not tell you
- "Decoded output is shown as text only; nothing was executed or rendered."
- Suggestions are guesses from the shape of the data. A string of letters and digits can match Base64 by accident.
- For a JWT: "Base64url-decoding each part shows its header and claims, but nothing here checks its signature or validity." Use the JWT / SAML Inspector for that.
- For SAML: "Decoding shows its XML; nothing here validates its signature or conditions."
- Limits: 5 MB input, 30 steps, 10 MB output per step. On screen, text is shown up to the first 256 KB and hex up to the first 4 KB.
Online activity and privacy
Nothing goes online. Input and results stay in memory and are not saved unless you export or add to a project.
Export and projects
The Export menu appears next to Copy result when the whole chain succeeded. It offers HTML, PDF, Markdown, Plain text, Redacted... and Add to project. See Export and redaction and Projects and notes.
The report lists the steps (number, name, ok or failed, bytes out, and any note or error, starting with the input size) and the first 4 KB of the final output as text, or the first 64 lines of its hex dump for binary output. The original input is not included separately.
Platform notes
No differences between macOS, Windows and Linux.
Troubleshooting
"Stopped here: not valid Base64, or larger than 10 MB when decoded". The data at that step is not in that encoding (or decodes to more than 10 MB). Check the step order, or try Base64url.
"Stopped here: bad percent escape at byte N". URL decode found a % not followed by two hex digits.
"input is not valid hex". Untick input is hex bytes, or remove non-hex characters.
"input larger than 5 MB" or "more than 30 steps". Reduce the input or the chain.
"output larger than 10 MB". A step (usually decompression) produced too much data and was stopped.
Copied result is shorter than expected. Very large text results are displayed, and copied, up to the first 256 KB.
Related tools
- Script Decoder for automatic decoding of obfuscated scripts.
- JWT / SAML Inspector to read and check tokens and SAML messages.
- Hash Calculator to hash a decoded value.
- IOC Extractor to pull indicators from decoded text.