Gzip Compress and Base64 Encode

How to gzip and Base64 encode

1. Type or paste text into the input, or drop a file. Live is on, so the output updates as you type.
2. Keep Format on gzip. zlib, raw deflate and, where the browser supports it, Brotli are in the same list.
3. Tick URL-safe (- _) if the string goes into a URL or a filename, or Wrap at 76 for email and PEM-style line lengths.
4. Copy output, Download the Base64 as a .txt file, or Download compressed file to save the binary .gz.

The meta line shows input size, compressed size, Base64 size and ratio, and the status says how much shorter or longer the result is than plain Base64 of the same input.

When compressing first is worth it

Base64 always grows data by a third. On data with repetition, gzip wins that back and more.

The figures were produced with gzip -6. The browser’s compressor can differ by a few bytes, and the meta line shows the exact numbers for your input. As a rule of thumb, inputs under about 100 bytes and anything already compressed are better off as plain Base64.

Where gzip + Base64 strings are used

• Cloud-init and VM user data: cloud-init detects gzip-compressed user data, which helps stay under the size limits of cloud metadata services.
• AWS CloudWatch Logs subscriptions deliver log batches to Lambda as Base64 of gzip in awslogs.data.
• Helm 3 stores each release as Base64 of gzipped JSON inside a Kubernetes Secret.
• Environment variables and JSON fields that must hold a config file or a large text blob.
• Query strings and deep links with a state object. Use URL-safe output there.

Apart from the H4sI prefix, nothing in the Base64 tells the receiver the data is gzip. Say so in the field name or the API docs, or, over HTTP, send the binary body with Content-Encoding: gzip instead of Base64.

Gzip + Base64 in code

import base64, gzip

data = open("products.json", "rb").read()
packed = base64.b64encode(gzip.compress(data, mtime=0)).decode()

// Node.js
const zlib = require("zlib");
const packed = zlib.gzipSync(Buffer.from(text)).toString("base64");

In browsers, new CompressionStream("gzip") does the compression, which is what this page uses. More examples are on the Python and Node.js pages.

Common problems

• The receiver gets garbage: it decoded the Base64 but did not gunzip. Check for H4sI and decompress with the gzip Base64 decoder.
• “incorrect header check” on the other side: the receiver expects zlib, not gzip. Switch Format to zlib (deflate); the zlib page explains which library expects which.
• Brotli is grayed out: the browser’s built-in compression API does not offer Brotli, and the page says so below the options.
• Output differs from another tool: gzip headers carry a timestamp and an OS byte, and compressors choose different levels. Different bytes can decompress to the same content.

FAQ