@fastify/compress
Adds compression utils to the Fastify reply object and a hook to decompress requests payloads.
Supports gzip, deflate, brotli, and zstd (Node.js 22.15+/23.8+).
Note: In large-scale scenarios, use a proxy like Nginx to handle response compression.
Warning: Since
@fastify/compressversion 4.x, payloads compressed with thezipalgorithm are not automatically uncompressed. This plugin focuses on response compression, andzipis not in the IANA Table of Content Encodings.
Install
npm i @fastify/compress
Compatibility
| Plugin version | Fastify version |
|---|---|
>=8.x |
^5.x |
>=6.x <8.x |
^4.x |
>=3.x <6.x |
^3.x |
^2.x |
^2.x |
>=0.x <2.x |
^1.x |
Please note that if a Fastify version is out of support, then so are the corresponding versions of this plugin in the table above. See Fastify's LTS policy for more details.
Usage - Compress replies
This plugin adds two functionalities to Fastify: a compress utility and a global compression hook.
Currently, the following encoding tokens are supported, using the first acceptable token in this order:
zstd(Node.js 22.15+/23.8+)brgzipdeflate*(no preference —@fastify/compresswill usegzip)identity(no compression)
If an unsupported encoding is received or the 'accept-encoding' header is missing, the payload will not be compressed.
To return an error for unsupported encoding, use the onUnsupportedEncoding option.
The plugin compresses payloads based on content-type. If absent, it assumes application/json.
Global hook
The global compression hook is enabled by default. To disable it, pass { global: false }:
await fastify.register(
import('@fastify/compress'),
{ global: false }
)
If only compression or decompression is required, set the globalCompression or globalDecompression config flags to false respectively (both are true by default).
await fastify.register(
import('@fastify/compress'),
// Disable compression but keep decompression enabled (default behavior for globalDecompression is true)
{ globalCompression: false }
)
// Disable decompression but keep compression enabled
await fastify.register(
import('@fastify/compress'),
{ globalDecompression: false }
)
Fastify encapsulation can be used to set global compression but run it only in a subset of routes by wrapping them inside a plugin.
Note: If using
@fastify/compressplugin together with@fastify/staticplugin,@fastify/compressmust be registered (with global hook) before registering@fastify/static.
Per Route options
Different compression options can be specified per route using the compress options in the route's configuration.
Setting compress: false on any route will disable compression on the route even if global compression is enabled.
await fastify.register(
import('@fastify/compress'),
{ global: false }
)
// only compress if the payload is above a certain size and use brotli
fastify.get('/custom-route', {
compress: {
inflateIfDeflated: true,
threshold: 128,
zlib: {
createBrotliCompress: () => createYourCustomBrotliCompress(),
createGzip: () => createYourCustomGzip(),
createDeflate: () => createYourCustomDeflate()
}
}, (req, reply) => {
// ...
})
reply.compress
This plugin adds a compress method to reply that compresses a stream or string based on the accept-encoding header. If a JS object is passed, it will be stringified to JSON.
The compress method uses per-route parameters if configured, otherwise it uses global parameters.
import fs from 'node:fs'
import fastify from 'fastify'
const app = fastify()
await app.register(import('@fastify/compress'), { global: false })
app.get('/', (req, reply) => {
reply
.type('text/plain')
.compress(fs.createReadStream('./package.json'))
})
await app.listen({ port: 3000 })
It's also possible to pass a Fetch API Response object or a Web ReadableStream. The plugin will automatically extract the body stream from the Response or convert the Web stream to a Node.js Readable behind the scenes.
import fastify from 'fastify'
const app = fastify()
await app.register(import('@fastify/compress'), { global: true })
app.get('/', async (req, reply) => {
const resp = new Response('Hello from Fetch Response')
reply.compress(resp)
})
app.get('/', async (req, reply) => {
return new Response('Hello from Fetch Response')
})
app.get('/', (req, reply) => {
const stream = new ReadableStream({
start (controller) {
controller.enqueue(new TextEncoder().encode('Hello from Web ReadableStream'))
controller.close()
}
})
reply.header('content-type', 'text/plain')
reply.compress(stream)
})
Compress Options
threshold
The minimum byte size for response compression. Defaults to 1024.
await fastify.register(
import('@fastify/compress'),
{ threshold: 2048 }
)
syncThreshold
The maximum byte size for compressing a response synchronously. Defaults to a value derived
from os.availableParallelism(), between 4096 and 65536.
Payloads that are already fully in memory (a string or a Buffer) and are not larger than
syncThreshold are compressed in one call instead of being pushed through a compression
stream. This is faster for small responses — a per-response zlib stream costs more in setup
and scheduling than the compression itself at these sizes — and it uses much less memory
while the response is in flight, since no zlib context is allocated and the source payload
is released immediately.
Larger payloads, streams, web ReadableStreams and fetch
Responses always use the streaming path, so the event loop is never held for long. Set
syncThreshold to 0 to compress every response through the streaming path.
await fastify.register(
import('@fastify/compress'),
{ syncThreshold: 65536 }
)
How the default is chosen
The streaming path hands compression to libuv's threadpool, so it only pays off when there are spare cores to overlap that work with the event loop. The more cores are available for that overlap, the smaller the payload at which streaming starts to win, so the default shrinks as the host gets wider:
| cores reported | default |
|---|---|
| 1 | 65536 |
| 2 | 8192 |
| 3 or more | 4096 |
The scale stops at four because libuv's threadpool defaults to four threads: beyond that,
extra cores cannot compress any more responses in parallel. The value is computed once at
startup and is always overridden by an explicit syncThreshold.
The memory saving does not depend on payload size — it comes from not allocating a zlib context per in-flight response — so it is retained in full even at the lowest default.
Raising it on CPU-constrained deployments
Since libuv 1.49 (Node.js 22.12), os.availableParallelism() accounts for a cgroup CPU
quota, so a container limited to whole cores reports those rather than the host's and gets
an appropriate default on its own. Two cases still read as a wide host and so get the
smallest default:
| deployment | reported | why |
|---|---|---|
| Node.js before 22.12 | host cores | the quota is not consulted at all |
a quota below one core, e.g. Kubernetes cpu: 500m |
host cores | the quota floors to zero and falls back to the host count |
Both are precisely the deployments that benefit most from compressing synchronously: with
little or no spare CPU, there is nothing for the streaming path to overlap with. Quotas are
also floored rather than rounded, so cpu: 2500m reports two cores.
If either applies to you, set syncThreshold explicitly:
await fastify.register(
import('@fastify/compress'),
// a container pinned to ~1 CPU behaves like a single core host
{ syncThreshold: 65536 }
)
The default is biased towards the low end on purpose: choosing too low a value only forgoes part of the possible throughput gain, while choosing too high a value costs considerably more than it can ever return.
Because the compressed size is known upfront, responses compressed synchronously carry an
accurate Content-Length instead of being sent with chunked transfer encoding. Replies that
already carry a Content-Length the plugin has been asked to keep
(see removeContentLengthHeader)
use the streaming path so the provided value is preserved.
customTypes
mime-db determines if a content-type should be compressed. Additional content types can be compressed via regex or a function.
await fastify.register(
import('@fastify/compress'),
{ customTypes: /x-protobuf$/ }
)
or
await fastify.register(
import('@fastify/compress'),
{ customTypes: contentType => contentType.endsWith('x-protobuf') }
)
onUnsupportedEncoding
Set onUnsupportedEncoding(encoding, request, reply) to send a custom error response for unsupported encoding. The function can modify the reply and return a string | Buffer | Stream | Error payload.
await fastify.register(
import('@fastify/compress'),
{
onUnsupportedEncoding: (encoding, request, reply) => {
reply.code(406)
return 'We do not support the ' + encoding + ' encoding.'
}
}
)
Disable compression by header
Response compression can be disabled by an x-no-compression header in the request.
Inflate pre-compressed bodies for clients that do not support compression
Optional feature to inflate pre-compressed data if the client does not include one of the supported compression types in its accept-encoding header.
await fastify.register(
import('@fastify/compress'),
{ inflateIfDeflated: true }
)
fastify.get('/file', (req, reply) =>
// will inflate the file on the way out for clients
// that indicate they do not support compression
reply.send(fs.createReadStream('./file.gz')))
Customize encoding priority
By default, @fastify/compress prioritizes compression as described here. Change this by passing an array of compression tokens to the encodings option:
await fastify.register(
import('@fastify/compress'),
// Only support gzip and deflate, and prefer deflate to gzip
{ encodings: ['deflate', 'gzip'] }
)
// Example with zstd support (Node.js 22.15+/23.8+)
await fastify.register(
import('@fastify/compress'),
// Prefer zstd, fallback to brotli, then gzip
{ encodings: ['zstd', 'br', 'gzip'] }
)
brotliOptions and zlibOptions
Compression can be tuned with brotliOptions and zlibOptions, which are passed directly to native node zlib methods. See class definitions.
server.register(fastifyCompress, {
brotliOptions: {
params: {
[zlib.constants.BROTLI_PARAM_MODE]: zlib.constants.BROTLI_MODE_TEXT, // useful for APIs that primarily return text
[zlib.constants.BROTLI_PARAM_QUALITY]: 4, // default is 4, max is 11, min is 0
},
},
zlibOptions: {
level: 6, // default is typically 6, max is 9, min is 0
}
});
Manage Content-Length header removal with removeContentLengthHeader
By default, @fastify/compress removes the reply Content-Length header. Change this by setting removeContentLengthHeader to false globally or per route.
// Global plugin scope
await server.register(fastifyCompress, { global: true, removeContentLengthHeader: false });
// Route-specific scope
fastify.get('/file', {
compress: { removeContentLengthHeader: false }
}, (req, reply) =>
reply.compress(fs.createReadStream('./file.gz'))
)
This option exists because the streaming path cannot know the compressed size upfront, so
the reply Content-Length would otherwise describe the uncompressed payload. Responses
taking the synchronous path do know the compressed size and always send an
accurate Content-Length; a reply that already carries a Content-Length and sets
removeContentLengthHeader to false is streamed instead, so the provided value is kept.
Usage - Decompress request payloads
This plugin adds a preParsing hook to decompress the request payload based on the content-encoding request header.
Currently, the following encoding tokens are supported:
zstd(Node.js 22.15+/23.8+)brgzipdeflate
If an unsupported encoding or invalid payload is received, the plugin throws an error.
If the request header is missing, the plugin yields to the next hook.
Supported payload types
The plugin supports compressing the following payload types:
- Strings and Buffers
- Node.js streams
- Response objects (from the Fetch API)
- ReadableStream objects (from the Web Streams API)
Global hook
The global request decompression hook is enabled by default. To disable it, pass { global: false }:
await fastify.register(
import('@fastify/compress'),
{ global: false }
)
Fastify encapsulation can be used to set global decompression but run it only in a subset of routes by wrapping them inside a plugin.
Per Route options
Specify different decompression options per route using the decompress options in the route's configuration.
await fastify.register(
import('@fastify/compress'),
{ global: false }
)
// Always decompress using gzip
fastify.get('/custom-route', {
decompress: {
forceRequestEncoding: 'gzip',
zlib: {
createBrotliDecompress: () => createYourCustomBrotliDecompress(),
createGunzip: () => createYourCustomGunzip(),
createInflate: () => createYourCustomInflate()
}
}
}, (req, reply) => {
// ...
})
requestEncodings
By default, @fastify/compress accepts all encodings specified here. Change this by passing an array of compression tokens to the requestEncodings option:
await fastify.register(
import('@fastify/compress'),
// Only support gzip
{ requestEncodings: ['gzip'] }
)
// Example with zstd support for request decompression (Node.js 22.15+/23.8+)
await fastify.register(
import('@fastify/compress'),
// Support zstd, brotli and gzip for request decompression
{ requestEncodings: ['zstd', 'br', 'gzip'] }
)
forceRequestEncoding
By default, @fastify/compress chooses the decompression algorithm based on the content-encoding header.
One algorithm can be forced, and the header ignored, by providing the forceRequestEncoding option.
If the request payload is not compressed, @fastify/compress will try to decompress, resulting in an error.
onUnsupportedRequestEncoding
The response error can be customized for unsupported request payload encoding by setting onUnsupportedEncoding(request, encoding) to a function that returns an error.
await fastify.register(
import('@fastify/compress'),
{
onUnsupportedRequestEncoding: (request, encoding) => {
return {
statusCode: 415,
code: 'UNSUPPORTED',
error: 'Unsupported Media Type',
message: 'We do not support the ' + encoding + ' encoding.'
}
}
}
)
onInvalidRequestPayload
The response error can be customized for undetectable request payloads by setting onInvalidRequestPayload(request, encoding) to a function that returns an error.
await fastify.register(
import('@fastify/compress'),
{
onInvalidRequestPayload: (request, encoding, error) => {
return {
statusCode: 400,
code: 'BAD_REQUEST',
error: 'Bad Request',
message: 'This is not a valid ' + encoding + ' encoded payload: ' + error.message
}
}
}
)
Acknowledgments
Past sp
- LetzDoIt
License
Licensed under MIT.