HTTPie CLI docs

Unstable version

If you want to try out the latest version of HTTPie that hasn't been officially released yet, you can install the development or unstable version directly from the master branch on GitHub. However, keep in mind that the development version is a work in progress and may not be as reliable as the stable version.

You can use the following command to install the development version of HTTPie on Linux, macOS, Windows, or FreeBSD operating systems. With this command, the code present in the master branch is downloaded and installed using pip.

$ python -m pip install --upgrade https://github.com/httpie/cli/archive/master.tar.gz

There are other ways to install the development version of HTTPie on macOS and Linux.

You can install it using Homebrew by running the following commands:

$ brew uninstall --force httpie
$ brew install --HEAD httpie

You can install it using Snapcraft by running the following commands:

$ snap remove httpie
$ snap install httpie --edge

To verify the installation, you can compare the version identifier on GitHub with the one available on your machine. You can check the version of HTTPie on your machine by using the command http --version.

$ http --version
# 3.X.X.dev0

Note that on your machine, the version name will have the .dev0 suffix.

Examples

Custom HTTP method, HTTP headers and JSON data:

$ http PUT pie.dev/put X-API-Token:123 name=John

Submitting forms:

$ http -f POST pie.dev/post hello=World

See the request that is being sent using one of the output options:

$ http -v pie.dev/get

Build and print a request without sending it using offline mode:

$ http --offline pie.dev/post hello=offline

Use GitHub API to post a comment on an issue with authentication:

$ http -a USERNAME POST https://api.github.com/repos/httpie/cli/issues/83/comments body='HTTPie is awesome! :heart:'

Upload a file using redirected input:

$ http pie.dev/post < files/data.json

Download a file and save it via redirected output:

$ http pie.dev/image/png > image.png

Download a file wget style:

$ http --download pie.dev/image/png

Use named sessions to make certain aspects of the communication persistent between requests to the same host:

$ http --session=logged-in -a username:password pie.dev/get API-Key:123

$ http --session=logged-in pie.dev/headers

Set a custom Host header to work around missing DNS records:

$ http localhost:8000 Host:example.com

Optional GET and POST

The METHOD argument is optional, and when you don’t specify it, HTTPie defaults to:

• GET for requests without body
• POST for requests with body

Here we don’t specify any request data, so both commands will send the same GET request:

$ http GET pie.dev/get

$ http pie.dev/get

Here, on the other hand, we do have some data, so both commands will make the same POST request:

$ http POST pie.dev/post hello=world

$ http pie.dev/post hello=world

Querystring parameters

If you find yourself manually constructing URLs with querystring parameters on the terminal, you may appreciate the param==value syntax for appending URL parameters.

With that, you don’t have to worry about escaping the & separators for your shell. Additionally, any special characters in the parameter name or value get automatically URL-escaped (as opposed to the parameters specified in the full URL, which HTTPie doesn’t modify).

$ http https://api.github.com/search/repositories q==httpie per_page==1

GET /search/repositories?q=httpie&per_page=1 HTTP/1.1

You can even retrieve the value from a file by using the param==@file syntax. This would also effectively strip the newlines from the end. See file based separators for more examples.

$ http pie.dev/get text==@files/text.txt

URL shortcuts for localhost

Additionally, curl-like shorthand for localhost is supported. This means that, for example, :3000 would expand to http://localhost:3000. If the port is omitted, then port 80 is assumed.

$ http :/foo

GET /foo HTTP/1.1
Host: localhost

$ http :3000/bar

GET /bar HTTP/1.1
Host: localhost:3000

$ http :

GET / HTTP/1.1
Host: localhost

Other default schemes

When HTTPie is invoked as https then the default scheme is https:// ($ https example.org will make a request to https://example.org).

You can also use the --default-scheme <URL_SCHEME> option to create shortcuts for other protocols than HTTP (possibly supported via plugins). Example for the httpie-unixsocket plugin:

# Before
$ http http+unix://%2Fvar%2Frun%2Fdocker.sock/info

# Create an alias
$ alias http-unix='http --default-scheme="http+unix"'

# Now the scheme can be omitted
$ http-unix %2Fvar%2Frun%2Fdocker.sock/info

--path-as-is

The standard behavior of HTTP clients is to normalize the path portion of URLs by squashing dot segments as a typical filesystem would:

$ http -v example.org/./../../etc/password

GET /etc/password HTTP/1.1

The --path-as-is option allows you to disable this behavior:

$ http --path-as-is -v example.org/./../../etc/password

GET /../../etc/password HTTP/1.1

File based separators

Using file contents as values for specific fields is a very common use case, which can be achieved through adding the @ suffix to the operators above. For example, instead of using a static string as the value for some header, you can use :@ operator to pass the desired value from a file.

$ http POST pie.dev/post \
    X-Data:@files/text.txt             # Read a header from a file
    token==@files/text.txt             # Read a query parameter from a file
    name=@files/text.txt               # Read a data field’s value from a file
    bookmarks:=@files/data.json        # Embed a JSON object from a file

Escaping rules

You can use \ to escape characters that shouldn’t be used as separators (or parts thereof). For instance, foo\==bar will become a data key/value pair (foo= and bar) instead of a URL parameter.

Often it is necessary to quote the values, e.g. foo='bar baz'.

If any of the field names or headers starts with a minus (e.g. -fieldname), you need to place all such items after the special token -- to prevent confusion with --arguments:

$ http pie.dev/post -- -name-starting-with-dash=foo -Unusual-Header:bar

POST /post HTTP/1.1
-Unusual-Header: bar
Content-Type: application/json

{
    "-name-starting-with-dash": "foo"
}

Default behavior

If your command includes some data request items, they are serialized as a JSON object by default. HTTPie also automatically sets the following headers, both of which can be overwritten:

Explicit JSON

You can use --json, -j to explicitly set Accept to application/json regardless of whether you are sending data (it’s a shortcut for setting the header via the usual header notation: http url Accept:'application/json, */*;q=0.5'). Additionally, HTTPie will try to detect JSON responses even when the Content-Type is incorrectly text/plain or unknown.

Non-string JSON fields

Non-string JSON fields use the := separator, which allows you to embed arbitrary JSON data into the resulting JSON object. Additionally, text and raw JSON files can also be embedded into fields using =@ and :=@:

$ http PUT pie.dev/put \
    name=John \                        # String (default)
    age:=29 \                          # Raw JSON — Number
    married:=false \                   # Raw JSON — Boolean
    hobbies:='["http", "pies"]' \      # Raw JSON — Array
    favorite:='{"tool": "HTTPie"}' \   # Raw JSON — Object
    bookmarks:=@files/data.json \      # Embed JSON file
    description=@files/text.txt        # Embed text file

PUT /person/1 HTTP/1.1
Accept: application/json, */*;q=0.5
Content-Type: application/json
Host: pie.dev

{
    "age": 29,
    "hobbies": [
        "http",
        "pies"
    ],
    "description": "John is a nice guy who likes pies.",
    "married": false,
    "name": "John",
    "favorite": {
        "tool": "HTTPie"
    },
    "bookmarks": {
        "HTTPie": "https://httpie.org",
    }
}

The :=/:=@ syntax is JSON-specific. You can switch your request to --form or --multipart, and string, float, and number values will continue to be serialized (as string form values). Other JSON types, however, are not allowed with --form or --multipart.

Nested JSON

If your use case involves sending complex JSON objects as part of the request body, HTTPie can help you build them right from your terminal. You still use the existing data field operators (=/:=) but instead of specifying a top-level field name (like key=value), you specify a path declaration. This tells HTTPie where and how to put the given value inside an object:

http pie.dev/post \
  platform[name]=HTTPie \
  platform[about][mission]='Make APIs simple and intuitive' \
  platform[about][homepage]=httpie.io \
  platform[about][homepage]=httpie.io \
  platform[about][stars]:=54000 \
  platform[apps][]=Terminal \
  platform[apps][]=Desktop \
  platform[apps][]=Web \
  platform[apps][]=Mobile

{
    "platform": {
        "name": "HTTPie",
        "about": {
            "mission": "Make APIs simple and intuitive",
            "homepage": "httpie.io",
            "stars": 54000
        },
        "apps": [
            "Terminal",
            "Desktop",
            "Web",
            "Mobile"
        ]
    }
}

$ http --offline --print=B pie.dev/post \
  category=tools \
  search[type]=id \
  search[id]:=1

{
    "category": "tools",
    "search": {
        "id": 1,
        "type": "id"
    }
}

$ http --offline --print=B pie.dev/post \
  category=tools \
  search[type]=keyword \
  search[keywords][]=APIs \
  search[keywords][]=CLI

{
    "category": "tools",
    "search": {
        "keywords": [
            "APIs",
            "CLI"
        ],
        "type": "keyword"
    }
}

$ http --offline --print=B pie.dev/post \
  category=tools \
  search[type]=keyword \
  search[keywords][1]=APIs \
  search[keywords][0]=CLI

{
    "category": "tools",
    "search": {
        "keywords": [
            "CLIs",
            "API"
        ],
        "type": "keyword"
    }
}

$ http --offline --print=B pie.dev/post \
  category=tools \
  search[type]=platforms \
  search[platforms][]=Terminal \
  search[platforms][1]=Desktop \
  search[platforms][3]=Mobile

{
    "category": "tools",
    "search": {
        "platforms": [
            "Terminal",
            "Desktop",
            null,
            "Mobile"
        ],
        "type": "platforms"
    }
}

$ http --offline --print=B pie.dev/post \
  category=tools \
  search[type]=platforms \
  'search[platforms]:=["Terminal", "Desktop"]' \
  search[platforms][]=Web \
  search[platforms][]=Mobile

{
    "category": "tools",
    "search": {
        "platforms": [
            "Terminal",
            "Desktop",
            "Web",
            "Mobile"
        ],
        "type": "platforms"
    }
}

$ http PUT pie.dev/put \
    shallow=value \                                # Shallow key-value pair
    object[key]=value \                            # Nested key-value pair
    array[]:=1 \                                   # Array — first item
    array[1]:=2 \                                  # Array — second item
    array[2]:=3 \                                  # Array — append (third item)
    very[nested][json][3][httpie][power][]=Amaze   # Nested object

$ http --offline --print=B pie.dev/post \
    []:=1 \
    []:=2 \
    []:=3

[
    1,
    2,
    3
]

http --offline --print=B pie.dev/post \
    [0][type]=platform [0][name]=terminal \
    [1][type]=platform [1][name]=desktop

[
    {
        "type": "platform",
        "name": "terminal"
    },
    {
        "type": "platform",
        "name": "desktop"
    }
]

$ http --offline --print=B pie.dev/post \
  'foo\[bar\]:=1' \
  'baz[\[]:=2' \
  'baz[\]]:=3'

{
    "baz": {
        "[": 2,
        "]": 3
    },
    "foo[bar]": 1
}

$ http --offline --print=B pie.dev/post \
  'backslash[\\]:=1'

{
    "backslash": {
        "\\": 1
    }
}

$ http --offline --print=B pie.dev/post \
  'object[\1]=stringified' \
  'object[\100]=same' \
  'array[1]=indexified'

{
    "array": [
        null,
        "indexified"
    ],
    "object": {
        "1": "stringified",
        "100": "same"
    }
}

$ http --offline --print=B pie.dev/post \
  'foo[bar]=OK' \
  'foo[baz][quux=FAIL'

HTTPie Syntax Error: Expecting ']'
foo[baz][quux
             ^

$ http --offline --print=B pie.dev/post \
  'array[]:=1' \
  'array[]:=2' \
  'array[key]:=3'
HTTPie Type Error: Can't perform 'key' based access on 'array' which has a type of 'array' but this operation requires a type of 'object'.
array[key]
     ^^^^^

$ http --offline --print=B pie.dev/post \
  user[name]:=411     # Defined as an integer
  user[name]=string   # Overridden with a string

{
    "user": {
        "name": "string"
    }
}

Raw JSON

For very complex JSON structures, it may be more convenient to pass it as raw request body, for example:

$ echo -n '{"hello": "world"}' | http POST pie.dev/post

$ http POST pie.dev/post < files/data.json

Regular forms

$ http --form POST pie.dev/post name='John Smith'

POST /post HTTP/1.1
Content-Type: application/x-www-form-urlencoded; charset=utf-8

name=John+Smith

File upload forms

If one or more file fields is present, the serialization and content type is multipart/form-data:

$ http -f POST pie.dev/post name='John Smith' cv@~/files/data.xml

The request above is the same as if the following HTML form were submitted:

<form enctype="multipart/form-data" method="post" action="http://example.com/jobs">
    <input type="text" name="name" />
    <input type="file" name="cv" />
</form>

Please note that @ is used to simulate a file upload form field, whereas =@ just embeds the file content as a regular text field value.

When uploading files, their content type is inferred from the file name. You can manually override the inferred content type:

$ http -f POST pie.dev/post name='John Smith' cv@'~/files/data.bin;type=application/pdf'

To perform a multipart/form-data request even without any files, use --multipart instead of --form:

$ http --multipart --offline example.org hello=world

POST / HTTP/1.1
Content-Length: 129
Content-Type: multipart/form-data; boundary=c31279ab254f40aeb06df32b433cbccb
Host: example.org

--c31279ab254f40aeb06df32b433cbccb
Content-Disposition: form-data; name="hello"

world
--c31279ab254f40aeb06df32b433cbccb--

File uploads are always streamed to avoid memory issues with large files.

By default, HTTPie uses a random unique string as the multipart boundary, but you can use --boundary to specify a custom string instead:

$ http --form --multipart --boundary=xoxo --offline example.org hello=world

POST / HTTP/1.1
Content-Length: 129
Content-Type: multipart/form-data; boundary=xoxo
Host: example.org

--xoxo
Content-Disposition: form-data; name="hello"

world
--xoxo--

If you specify a custom Content-Type header without including the boundary bit, HTTPie will add the boundary value (explicitly specified or auto-generated) to the header automatically:

$ http --form --multipart --offline example.org hello=world Content-Type:multipart/letter

POST / HTTP/1.1
Content-Length: 129
Content-Type: multipart/letter; boundary=c31279ab254f40aeb06df32b433cbccb
Host: example.org

--c31279ab254f40aeb06df32b433cbccb
Content-Disposition: form-data; name="hello"

world
--c31279ab254f40aeb06df32b433cbccb--

Default request headers

There are a couple of default headers that HTTPie sets:

GET / HTTP/1.1
Accept: */*
Accept-Encoding: gzip, deflate
User-Agent: HTTPie/<version>
Host: <taken-from-URL>

All of these can be overwritten or unset (see below).

Reading headers from a file

You can read headers from a file by using the :@ operator. This would also effectively strip the newlines from the end. See file based separators for more examples.

$ http pie.dev/headers X-Data:@files/text.txt

Empty headers and header un-setting

To unset a previously specified header (such a one of the default headers), use Header::

$ http pie.dev/headers Accept: User-Agent:

To send a header with an empty value, use Header;, with a semicolon:

$ http pie.dev/headers 'Header;'

Please note that some internal headers, such as Content-Length, can’t be unset if they are automatically added by the client itself.

Multiple header values with the same name

If the request is sent with multiple headers that are sharing the same name, then the HTTPie will send them individually.

http --offline example.org Cookie:one Cookie:two

GET / HTTP/1.1
Cookie: one
Cookie: two

It is also possible to pass a single header value pair, where the value is a comma separated list of header values. Then the client will send it as a single header.

http --offline example.org Numbers:one,two

GET / HTTP/1.1
Numbers: one,two

Also be aware that if the current session contains any headers they will get overwritten by individual commands when sending a request instead of being joined together.

Limiting response headers

The --max-headers=n option allows you to control the number of headers HTTPie reads before giving up (the default 0, i.e., there’s no limit).

$ http --max-headers=100 pie.dev/get

Basic auth

$ http -a username:password pie.dev/basic-auth/username/password

Digest auth

$ http -A digest -a username:password pie.dev/digest-auth/httpie/username/password

Bearer auth

https -A bearer -a token pie.dev/bearer

Password prompt

If you omit the password part of --auth, -a, HTTPie securely prompts you for it:

$ http -a username pie.dev/basic-auth/username/password

Please note that when you use --session, prompted passwords are persisted in session files.

Empty password

To send an empty password without being prompted for it, include a trailing colon in the credentials:

$ http -a username: pie.dev/headers

.netrc

Authentication information from your ~/.netrc file is by default honored as well.

For example:

$ cat ~/.netrc
machine pie.dev
login httpie
password test

$ http pie.dev/basic-auth/httpie/test
HTTP/1.1 200 OK
[...]

This can be disabled with the --ignore-netrc option:

$ http --ignore-netrc pie.dev/basic-auth/httpie/test
HTTP/1.1 401 UNAUTHORIZED
[...]

Auth plugins

Additional authentication mechanism can be installed as plugins. They can be found on the Python Package Index. Here are a few picks:

• httpie-api-auth: ApiAuth
• httpie-aws-auth: AWS / Amazon S3
• httpie-edgegrid: EdgeGrid
• httpie-hmac-auth: HMAC
• httpie-jwt-auth: JWTAuth (JSON Web Tokens)
• httpie-negotiate: SPNEGO (GSS Negotiate)
• httpie-ntlm: NTLM (NT LAN Manager)
• httpie-oauth1: OAuth 1.0a
• requests-hawk: Hawk

See plugin manager for more details.

Follow Location

To instruct HTTPie to follow the Location header of 30x responses and show the final response instead, use the --follow, -F option:

$ http --follow pie.dev/redirect/3

With 307 Temporary Redirect and 308 Permanent Redirect, the method and the body of the original request are reused to perform the redirected request. Otherwise, a body-less GET request is performed.

Showing intermediary redirect responses

If you wish to see the intermediary requests/responses, then use the --all option:

$ http --follow --all pie.dev/redirect/3

Limiting maximum redirects followed

To change the default limit of maximum 30 redirects, use the --max-redirects=<limit> option:

$ http --follow --all --max-redirects=2 pie.dev/redirect/3

Environment variables

You can also configure proxies by environment variables ALL_PROXY, HTTP_PROXY and HTTPS_PROXY, and the underlying Requests library will pick them up. If you want to disable proxies configured through the environment variables for certain hosts, you can specify them in NO_PROXY.

In your ~/.bash_profile:

export HTTP_PROXY=http://10.10.1.10:3128
export HTTPS_PROXY=https://10.10.1.10:1080
export NO_PROXY=localhost,example.com

SOCKS

Usage for SOCKS is the same as for other types of proxies:

$ http --proxy=http:socks5://user:pass@host:port --proxy=https:socks5://user:pass@host:port example.org

Server SSL certificate verification

To skip the host’s SSL certificate verification, you can pass --verify=no (default is yes):

$ http --verify=no https://pie.dev/get

Custom CA bundle

You can also use --verify=<CA_BUNDLE_PATH> to set a custom CA bundle path:

$ http --verify=/ssl/custom_ca_bundle https://example.org

Client side SSL certificate

To use a client side certificate for the SSL communication, you can pass the path of the cert file with --cert:

$ http --cert=client.pem https://example.org

If the private key is not contained in the cert file, you may pass the path of the key file with --cert-key:

$ http --cert=client.crt --cert-key=client.key https://example.org

If the given private key requires a passphrase, HTTPie will automatically detect it and ask it through a prompt:

$ http --cert=client.pem --cert-key=client.key https://example.org
http: passphrase for client.key: ****

If you don't want to see a prompt, you can supply the passphrase with the --cert-key-pass argument:

$ http --cert=client.pem --cert-key=client.key --cert-key-pass=my_password https://example.org

SSL version

Use the --ssl=<PROTOCOL> option to specify the desired protocol version to use. This will default to SSL v2.3 which will negotiate the highest protocol that both the server and your installation of OpenSSL support. The available protocols are ssl2.3, ssl3, tls1, tls1.1, tls1.2, tls1.3. (The actually available set of protocols may vary depending on your OpenSSL installation.)

# Specify the vulnerable SSL v3 protocol to talk to an outdated server:
$ http --ssl=ssl3 https://vulnerable.example.org

SSL ciphers

You can specify the available ciphers with --ciphers. It should be a string in the OpenSSL cipher list format.

$ http --ciphers=ECDHE-RSA-AES128-GCM-SHA256 https://pie.dev/get

Note: these cipher strings do not change the negotiated version of SSL or TLS, they only affect the list of available cipher suites.

To see the default cipher string, run http --help and see the --ciphers section under SSL.

What parts of the HTTP exchange should be printed

All the other output options are under the hood just shortcuts for the more powerful --print, -p. It accepts a string of characters each of which represents a specific part of the HTTP exchange:

Print request and response headers:

$ http --print=Hh PUT pie.dev/put hello=world

$ http --meta pie.dev/delay/1

Elapsed time: 1.099171542s

$ http --print=hm pie.dev/get

HTTP/1.1 200 OK
Content-Type: application/json

Elapsed time: 0.077538375s

Verbose output

--verbose can often be useful for debugging the request and generating documentation examples:

$ http --verbose PUT pie.dev/put hello=world
PUT /put HTTP/1.1
Accept: application/json, */*;q=0.5
Accept-Encoding: gzip, deflate
Content-Type: application/json
Host: pie.dev
User-Agent: HTTPie/0.2.7dev

{
    "hello": "world"
}

HTTP/1.1 200 OK
Connection: keep-alive
Content-Length: 477
Content-Type: application/json
Date: Sun, 05 Aug 2012 00:25:23 GMT
Server: gunicorn/0.13.4

{
    […]
}

# Just like the above, but with additional columns like the total elapsed time
$ http -vv pie.dev/get

Quiet output

--quiet redirects all output that would otherwise go to stdout and stderr to /dev/null (except for errors and warnings). This doesn’t affect output to a file via --output or --download.

# There will be no output:
$ http --quiet pie.dev/post enjoy='the silence'

If you’d like to silence warnings as well, use -q or --quiet twice:

# There will be no output, even in case of an unexpected response status code:
$ http -qq --check-status pie.dev/post enjoy='the silence without warnings'

Update warnings

When there is a new release available for your platform (for example; if you installed HTTPie through pip, it will check the latest version on PyPI), HTTPie will regularly warn you about the new update (once a week). If you want to disable this behavior, you can set disable_update_warnings to true in your config file.

Viewing intermediary requests/responses

To see all the HTTP communication, i.e. the final request/response as well as any possible intermediary requests/responses, use the --all option. The intermediary HTTP communication include followed redirects (with --follow), the first unauthorized request when HTTP digest authentication is used (--auth=digest), etc.

# Include all responses that lead to the final one:
$ http --all --follow pie.dev/redirect/3

The intermediary requests/responses are by default formatted according to --print, -p (and its shortcuts described above).

Conditional body download

As an optimization, the response body is downloaded from the server only if it’s part of the output. This is similar to performing a HEAD request, except that it applies to any HTTP method you use.

Let’s say that there is an API that returns the whole resource when it is updated, but you are only interested in the response headers to see the status code after an update:

$ http --headers PATCH pie.dev/patch name='New Name'

Since you are only printing the HTTP headers here, the connection to the server is closed as soon as all the response headers have been received. Therefore, bandwidth and time isn’t wasted downloading the body which you don’t care about. The response headers are downloaded always, even if they are not part of the output

Redirected Input

The universal method for passing request data is through redirected stdin (standard input)—piping.

By default, stdin data is buffered and then with no further processing used as the request body. If you provide Content-Length, then the request body is streamed without buffering. You may also use --chunked to enable streaming via chunked transfer encoding or --compress, -x to compress the request body.

There are multiple useful ways to use piping:

Redirect from a file:

$ http PUT pie.dev/put X-API-Token:123 < files/data.json

Or the output of another program:

$ grep '401 Unauthorized' /var/log/httpd/error_log | http POST pie.dev/post

You can use echo for simple data:

$ echo -n '{"name": "John"}' | http PATCH pie.dev/patch X-API-Token:123

You can also use a Bash here string:

$ http pie.dev/post <<<'{"name": "John"}'

You can even pipe web services together using HTTPie:

$ http GET https://api.github.com/repos/httpie/cli | http POST pie.dev/post

You can use cat to enter multiline data on the terminal:

$ cat | http POST pie.dev/post
<paste>
^D

$ cat | http POST pie.dev/post Content-Type:text/plain
- buy milk
- call parents
^D

On macOS, you can send the contents of the clipboard with pbpaste:

$ pbpaste | http PUT pie.dev/put

Passing data through stdin can’t be combined with data fields specified on the command line:

$ echo -n 'data' | http POST example.org more=data  # This is invalid

To prevent HTTPie from reading stdin data you can use the --ignore-stdin option.

Request data via --raw

In a situation when piping data via stdin is not convenient (for example, when generating API docs examples), you can specify the raw request body via the --raw option.

$ http --raw 'Hello, world!' pie.dev/post

$ http --raw '{"name": "John"}' pie.dev/post

Request data from a filename

An alternative to redirected stdin is specifying a filename (as @/path/to/file) whose content is used as if it came from stdin.

It has the advantage that the Content-Type header is automatically set to the appropriate value based on the filename extension. For example, the following request sends the verbatim contents of that XML file with Content-Type: application/xml:

$ http PUT pie.dev/put @files/data.xml

File uploads are always streamed to avoid memory issues with large files.

Colors and formatting

Syntax highlighting is applied to HTTP headers and bodies (where it makes sense). You can choose your preferred color scheme via the --style option if you don’t like the default one. There are dozens of styles available, here are just a few notable ones:

Use one of these options to control output processing:

HTTPie looks at Content-Type to select the right syntax highlighter and formatter for each message body. If that fails (e.g., the server provides the wrong type), or you prefer a different treatment, you can manually overwrite the mime type for a response with --response-mime:

$ http --response-mime=text/yaml pie.dev/get

Formatting has the following effects:

• HTTP headers are sorted by name.
• JSON data is indented, sorted by keys, and unicode escapes are converted to the characters they represent.
• XML and XHTML data is indented.

Please note that sometimes there might be changes made by formatters on the actual response body (e.g., collapsing empty tags on XML) but the end result will always be semantically indistinguishable. Some of these formatting changes can be configured more granularly through format options.

Format options

The --format-options=opt1:value,opt2:value option allows you to control how the output should be formatted when formatting is applied. The following options are available:

For example, this is how you would disable the default header and JSON key sorting, and specify a custom JSON indent size:

$ http --format-options headers.sort:false,json.sort_keys:false,json.indent:2 pie.dev/get

There are also two shortcuts that allow you to quickly disable and re-enable sorting-related format options (currently it means JSON keys and headers): --unsorted and --sorted.

This is something you will typically store as one of the default options in your config file.

Redirected output

HTTPie uses a different set of defaults for redirected output than for terminal output. The differences being:

• Formatting and colors aren’t applied (unless --pretty is specified).
• Only the response body is printed (unless one of the output options is set).
• Also, binary data isn’t suppressed.

The reason is to make piping HTTPie’s output to another programs and downloading files work with no extra flags. Most of the time, only the raw response body is of an interest when the output is redirected.

Download a file:

$ http pie.dev/image/png > image.png

Download an image of an Octocat, resize it using ImageMagick, and upload it elsewhere:

$ http octodex.github.com/images/original.jpg | convert - -resize 25% - | http example.org/Octocats

Force colorizing and formatting, and show both the request and the response in less pager:

$ http --pretty=all --verbose pie.dev/get | less -R

The -R flag tells less to interpret color escape sequences included HTTPie’s output.

You can create a shortcut for invoking HTTPie with colorized and paged output by adding the following to your ~/.bash_profile:

function httpless {
    # `httpless example.org'
    http --pretty=all --print=hb "$@" | less -R;
}

Binary data

Binary data is suppressed for terminal output, which makes it safe to perform requests to URLs that send back binary data. Binary data is also suppressed in redirected but prettified output. The connection is closed as soon as we know that the response body is binary,

$ http pie.dev/bytes/2000

You will nearly instantly see something like this:

HTTP/1.1 200 OK
Content-Type: application/octet-stream

+-----------------------------------------+
| NOTE: binary data not shown in terminal |
+-----------------------------------------+

Display encoding

HTTPie tries to do its best to decode message bodies when printing them to the terminal correctly. It uses the encoding specified in the Content-Type charset attribute. If a message doesn’t define its charset, we auto-detect it. For very short messages (1–32B), where auto-detection would be unreliable, we default to UTF-8. For cases when the response encoding is still incorrect, you can manually overwrite the response charset with --response-charset:

$ http --response-charset=big5 pie.dev/get

Downloaded filename

There are three mutually exclusive ways through which HTTPie determines the output filename (with decreasing priority):

1. You can explicitly provide it via --output, -o. The file gets overwritten if it already exists (or appended to with --continue, -c).
2. The server may specify the filename in the optional Content-Disposition response header. Any leading dots are stripped from a server-provided filename.
3. The last resort HTTPie uses is to generate the filename from a combination of the request URL and the response Content-Type. The initial URL is always used as the basis for the generated filename — even if there has been one or more redirects.

To prevent data loss by overwriting, HTTPie adds a unique numerical suffix to the filename when necessary (unless specified with --output, -o).

Piping while downloading

You can also redirect the response body to another program while the response headers and progress are still shown in the terminal:

$ http -d https://github.com/httpie/cli/archive/master.tar.gz | tar zxf -

Resuming downloads

If --output, -o is specified, you can resume a partial download using the --continue, -c option. This only works with servers that support Range requests and 206 Partial Content responses. If the server doesn’t support that, the whole file will simply be downloaded:

$ http -dco file.zip example.org/file

-dco is shorthand for --download --continue --output.

Other notes

• The --download option only changes how the response body is treated.
• You can still set custom headers, use sessions, --verbose, -v, etc.
• --download always implies --follow (redirects are followed).
• --download also implies --check-status (error HTTP status will result in a non-zero exist static code).
• HTTPie exits with status code 1 (error) if the body hasn’t been fully downloaded.
• Accept-Encoding can’t be set with --download.

Disabling buffering

You can use the --stream, -S flag to make two things happen:

1. The output is flushed in much smaller chunks without any buffering, which makes HTTPie behave kind of like tail -f for URLs.
2. Streaming becomes enabled even when the output is prettified: It will be applied to each line of the response and flushed immediately. This makes it possible to have a nice output for long-lived requests, such as one to the Twitter streaming API.

The --stream option is automatically enabled when the response headers include Content-Type: text/event-stream.

Example use cases

Prettified streamed response:

$ http --stream pie.dev/stream/3

Streamed output by small chunks à la tail -f:

# Send each new line (JSON object) to another URL as soon as it arrives from a streaming API:
$ http --stream pie.dev/stream/3 | while read line; do echo "$line" | http pie.dev/post ; done

Named sessions

You can create one or more named session per host. For example, this is how you can create a new session named user1 for pie.dev:

$ http --session=user1 -a user1:password pie.dev/get X-Foo:Bar

From now on, you can refer to the session by its name (user1). When you choose to use the session again, all previously specified authentication or HTTP headers will automatically be set:

$ http --session=user1 pie.dev/get

To create or reuse a different session, simply specify a different name:

$ http --session=user2 -a user2:password pie.dev/get X-Bar:Foo

Named sessions’ data is stored in JSON files inside the sessions subdirectory of the config directory, typically ~/.config/httpie/sessions/<host>/<name>.json (%APPDATA%\httpie\sessions\<host>\<name>.json on Windows).

If you have executed the above commands on a Unix machine, you should be able to list the generated sessions files using:

$ ls -l ~/.config/httpie/sessions/pie.dev

Anonymous sessions

Instead of giving it a name, you can also directly specify a path to a session file. This allows for sessions to be re-used across multiple hosts:

# Create a session:
$ http --session=/tmp/session.json example.org

# Use the session to make a request to another host:
$ http --session=/tmp/session.json admin.example.org

# You can also refer to a previously created named session:
$ http --session=~/.config/httpie/sessions/another.example.org/test.json example.org

When creating anonymous sessions, please remember to always include at least one /, even if the session files is located in the current directory (i.e. --session=./session.json instead of just --session=session.json), otherwise HTTPie assumes a named session instead.

Readonly session

To use the original session file without updating it from the request/response exchange after it has been created, specify the session name via --session-read-only=SESSION_NAME_OR_PATH instead.

# If the session file doesn’t exist, then it is created:
$ http --session-read-only=./ro-session.json pie.dev/headers Custom-Header:orig-value

# But it is not updated:
$ http --session-read-only=./ro-session.json pie.dev/headers Custom-Header:new-value

Host-based cookie policy

Cookies persisted in sessions files have a domain field. This binds them to a specified hostname. For example:

{
    "cookies": [
        {
            "domain": "pie.dev",
            "name": "pie",
            "value": "apple"
        },
        {
            "domain": "httpbin.org",
            "name": "bin",
            "value": "http"
        }
    ]
}

Using this session file, we include Cookie: pie=apple only in requests against pie.dev and subdomains (e.g., foo.pie.dev or foo.bar.pie.dev):

$ http --session=./session.json pie.dev/cookies

{
    "cookies": {
        "pie": "apple"
    }
}

To make a cookie domain unbound (i.e., to make it available to all hosts, including throughout a cross-domain redirect chain), you can set the domain field to null in the session file:

{
    "cookies": [
        {
            "domain": null,
            "name": "unbound-cookie",
            "value": "send-me-to-any-host"
        }
    ]
}

$ http --session=./session.json pie.dev/cookies

{
    "cookies": {
        "unbound-cookie": "send-me-to-any-host"
    }
}

Cookie storage behavior

There are three possible sources of persisted cookies within a session. They have the following storage priority: 1—response; 2—command line; 3—session file.

1. Receive a response with a Set-Cookie header:

   $ http --session=./session.json pie.dev/cookie/set?foo=bar

   Run

2. Send a cookie specified on the command line as seen in cookies:

   $ http --session=./session.json pie.dev/headers Cookie:foo=bar

   Run

3. Manually set cookie parameters in the session file:

   {
      "cookies": {
          "foo": {
              "expires": null,
              "path": "/",
              "secure": false,
              "value": "bar"
              }
      }
   }

In summary:

• Cookies set via the CLI overwrite cookies of the same name inside session files.
• Server-sent Set-Cookie header cookies overwrite any pre-existing ones with the same name.

Cookie expiration handling:

• When the server expires an existing cookie, HTTPie removes it from the session file.
• When a cookie in a session file expires, HTTPie removes it before sending a new request.

Upgrading sessions

HTTPie may introduce changes in the session file format. When HTTPie detects an obsolete format, it shows a warning. You can upgrade your session files using the following commands:

Upgrade all existing named sessions inside the sessions subfolder of your config directory:

$ httpie cli sessions upgrade-all
Upgraded 'api_auth' @ 'pie.dev' to v3.1.0
Upgraded 'login_cookies' @ 'httpie.io' to v3.1.0

Upgrading individual sessions requires you to specify the session's hostname. That allows HTTPie to find the correct file in the case of name sessions. Additionally, it allows it to correctly bind cookies when upgrading with --bind-cookies.

Upgrade a single named session:

$ httpie cli sessions upgrade pie.dev api_auth
Upgraded 'api_auth' @ 'pie.dev' to v3.1.0

Upgrade a single anonymous session using a file path:

$ httpie cli sessions upgrade pie.dev ./session.json
Upgraded 'session.json' @ 'pie.dev' to v3.1.0

Config file directory

To see the exact location for your installation, run http --debug and look for config_dir in the output.

The default location of the configuration file on most platforms is $XDG_CONFIG_HOME/httpie/config.json (defaulting to ~/.config/httpie/config.json).

For backward compatibility, if the directory ~/.httpie exists, the configuration file there will be used instead.

On Windows, the config file is located at %APPDATA%\httpie\config.json.

The config directory can be changed by setting the $HTTPIE_CONFIG_DIR environment variable:

$ export HTTPIE_CONFIG_DIR=/tmp/httpie
$ http pie.dev/get

Configurable options

Currently, HTTPie offers a single configurable option:

$ cat ~/.config/httpie/config.json

{
    "default_options": [
        "--style=fruity"
    ]
}

Un-setting previously specified options

Default options from the config file, or specified any other way, can be unset for a particular invocation via --no-OPTION arguments passed via the command line (e.g., --no-style or --no-session).

Best practices

The default behavior of automatically reading stdin is typically not desirable during non-interactive invocations. You most likely want to use the --ignore-stdin option to disable it.

It's important to note that without the --ignore-stdin option, HTTPie may appear to have stopped working (hang). This happens because, in situations where HTTPie is invoked outside of an interactive session, such as from a cron job, stdin is not connected to a terminal. This means that the rules for redirected input will be followed. When stdin is redirected, HTTPie assumes that the input will contain the request body, and it waits for the input to be provided. But, since there is neither any input data nor an end-of-file (EOF) signal, HTTPie gets stuck. To avoid this problem, the --ignore-stdin flag should be used in scripts, unless data is being piped to HTTPie.

To prevent your program from becoming unresponsive when the server fails to respond, it's a good idea to use the --timeout option to set a connection timeout limit.

httpie cli

$ httpie cli check-updates

$ httpie cli export-args | jq '"Program: " + .spec.name + ", Version: " +  .version'
"Program: http, Version: 0.0.1a0"

$ httpie cli plugins install httpie-plugin
Installing httpie-plugin...
Successfully installed httpie-plugin-1.0.2

$ httpie cli plugins list
httpie_plugin (1.0.2)
  httpie_plugin (httpie.plugins.auth.v1)
httpie_plugin_2 (1.0.6)
  httpie_plugin_2 (httpie.plugins.auth.v1)
httpie_converter (1.0.0)
  httpie_iterm_converter (httpie.plugins.converter.v1)
  httpie_konsole_konverter (httpie.plugins.converter.v1)

$ httpie cli plugins upgrade httpie-plugin

$ httpie cli plugins uninstall httpie-plugin

Interface design

The syntax of the command arguments closely correspond to the actual HTTP requests sent over the wire. It has the advantage that it’s easy to remember and read. You can often translate an HTTP request to an HTTPie argument list just by inlining the request elements. For example, compare this HTTP request:

POST /post HTTP/1.1
Host: pie.dev
X-API-Key: 123
User-Agent: Bacon/1.0
Content-Type: application/x-www-form-urlencoded

name=value&name2=value2

with the HTTPie command that sends it:

$ http -f POST pie.dev/post \
    X-API-Key:123 \
    User-Agent:Bacon/1.0 \
    name=value \
    name2=value2

Notice that both the order of elements and the syntax are very similar, and that only a small portion of the command is used to control HTTPie and doesn’t directly correspond to any part of the request (here, it’s only -f asking HTTPie to send a form request).

The two modes, --pretty=all (default for terminal) and --pretty=none (default for redirected output), allow for both user-friendly interactive use and usage from scripts, where HTTPie serves as a generic HTTP client.

In the future, the command line syntax and some of the --OPTIONS may change slightly, as HTTPie improves and new features are added. All changes are recorded in the change log.

Community and Support

HTTPie has the following community channels:

• GitHub Issues for bug reports and feature requests
• Discord server to ask questions, discuss features, and for general API development discussion
• StackOverflow to ask questions (make sure to use the httpie tag)

Related projects

Contributing

See CONTRIBUTING.

Security policy

See github.com/httpie/cli/security/policy.

Change log

See CHANGELOG.

Artwork

• README Animation by Allen Smith.

Licence

BSD-3-Clause: LICENSE.

Authors

Jakub Roztocil (@jakubroztocil) created HTTPie and these fine people have contributed.