#htmx in a Nutshell

htmx is a library that allows you to access modern browser features directly from HTML, rather than using javascript.

To understand htmx, first lets take a look at an anchor tag:

<a href="/blog">Blog</a>

This anchor tag tells a browser:

“When a user clicks on this link, issue an HTTP GET request to ‘/blog’ and load the response content into the browser window”.

With that in mind, consider the following bit of HTML:

<button hx-post="/clicked"
    Click Me!

This tells htmx:

“When a user clicks on this button, issue an HTTP POST request to ‘/clicked’ and use the content from the response to replace the element with the id parent-div in the DOM”

htmx extends and generalizes the core idea of HTML as a hypertext, opening up many more possibilities directly within the language:

  • Now any element, not just anchors and forms, can issue an HTTP request
  • Now any event, not just clicks or form submissions, can trigger requests
  • Now any HTTP verb, not just GET and POST, can be used
  • Now any element, not just the entire window, can be the target for update by the request

Note that when you are using htmx, on the server side you typically respond with HTML, not JSON. This keeps you firmly within the original web programming model, using Hypertext As The Engine Of Application State without even needing to really understand that concept.

It’s worth mentioning that, if you prefer, you can use the data- prefix when using htmx:

<a data-hx-post="/click">Click Me!</a>


Htmx is a dependency-free, browser-oriented javascript library. This means that using it is as simple as adding a <script> tag to your document head. No need for complicated build steps or systems.

If you are migrating to htmx from intercooler.js, please see the migration guide.

#Via A CDN (e.g.

The fastest way to get going with htmx is to load it via a CDN. You can simply add this to your head tag and get going:

<script src="" integrity="sha384-ujb1lZYygJmzgSwoxRggbCHcjc0rB2XoQrxeTUQyRjrOnlCoYta87iKBWq3EsdM2" crossorigin="anonymous"></script>

While the CDN approach is extremely simple, you may want to consider not using CDNs in production.

#Download a copy

The next easiest way to install htmx is to simply copy it into your project.

Download htmx.min.js from and add it to the appropriate directory in your project and include it where necessary with a <script> tag:

<script src="/path/to/htmx.min.js"></script>

You can also add extensions this way, by downloading them from the ext/ directory.


For npm-style build systems, you can install htmx via npm:

npm install

After installing, you’ll need to use appropriate tooling to use node_modules/ (or .min.js). For example, you might bundle htmx with some extensions and project-specific code.


If you are using webpack to manage your javascript:

  • Install htmx via your favourite package manager (like npm or yarn)
  • Add the import to your index.js
import '';

If you want to use the global htmx variable (recommended), you need to inject it to the window scope:

  • Create a custom JS file
  • Import this file to your index.js (below the import from step 2)
import 'path/to/my_custom.js';
  • Then add this code to the file:
window.htmx = require('');
  • Finally, rebuild your bundle


The core of htmx is a set of attributes that allow you to issue AJAX requests directly from HTML:

hx-getIssues a GET request to the given URL
hx-postIssues a POST request to the given URL
hx-putIssues a PUT request to the given URL
hx-patchIssues a PATCH request to the given URL
hx-deleteIssues a DELETE request to the given URL

Each of these attributes takes a URL to issue an AJAX request to. The element will issue a request of the specified type to the given URL when the element is triggered:

<div hx-put="/messages">
    Put To Messages

This tells the browser:

When a user clicks on this div, issue a PUT request to the URL /messages and load the response into the div

#Triggering Requests

By default, AJAX requests are triggered by the “natural” event of an element:

  • input, textarea & select are triggered on the change event
  • form is triggered on the submit event
  • everything else is triggered by the click event

If you want different behavior you can use the hx-trigger attribute to specify which event will cause the request.

Here is a div that posts to /mouse_entered when a mouse enters it:

<div hx-post="/mouse_entered" hx-trigger="mouseenter">
    [Here Mouse, Mouse!]

#Trigger Modifiers

A trigger can also have a few additional modifiers that change its behavior. For example, if you want a request to only happen once, you can use the once modifier for the trigger:

<div hx-post="/mouse_entered" hx-trigger="mouseenter once">
    [Here Mouse, Mouse!]

Other modifiers you can use for triggers are:

  • changed - only issue a request if the value of the element has changed
  • delay:<time interval> - wait the given amount of time (e.g. 1s) before issuing the request. If the event triggers again, the countdown is reset.
  • throttle:<time interval> - wait the given amount of time (e.g. 1s) before issuing the request. Unlike delay if a new event occurs before the time limit is hit the event will be discarded, so the request will trigger at the end of the time period.
  • from:<CSS Selector> - listen for the event on a different element. This can be used for things like keyboard shortcuts.

You can use these attributes to implement many common UX patterns, such as Active Search:

<input type="text" name="q"
    hx-trigger="keyup changed delay:500ms"
<div id="search-results"></div>

This input will issue a request 500 milliseconds after a key up event if the input has been changed and inserts the results into the div with the id search-results.

Multiple triggers can be specified in the hx-trigger attribute, separated by commas.

#Trigger Filters

You may also apply trigger filters by using square brackets after the event name, enclosing a javascript expression that will be evaluated. If the expression evaluates to true the event will trigger, otherwise it will not.

Here is an example that triggers only on a Control-Click of the element

<div hx-get="/clicked" hx-trigger="click[ctrlKey]">
    Control Click Me

Properties like ctrlKey will be resolved against the triggering event first, then the global scope. The this symbol will be set to the current element.

#Special Events

htmx provides a few special events for use in hx-trigger:

  • load - fires once when the element is first loaded
  • revealed - fires once when an element first scrolls into the viewport
  • intersect - fires once when an element first intersects the viewport. This supports two additional options:
    • root:<selector> - a CSS selector of the root element for intersection
    • threshold:<float> - a floating point number between 0.0 and 1.0, indicating what amount of intersection to fire the event on

You can also use custom events to trigger requests if you have an advanced use case.


If you want an element to poll the given URL rather than wait for an event, you can use the every syntax with the hx-trigger attribute:

<div hx-get="/news" hx-trigger="every 2s"></div>

This tells htmx

Every 2 seconds, issue a GET to /news and load the response into the div

If you want to stop polling from a server response you can respond with the HTTP response code 286 and the element will cancel the polling.

#Load Polling

Another technique that can be used to achieve polling in htmx is “load polling”, where an element specifies a load trigger along with a delay, and replaces itself with the response:

<div hx-get="/messages"
    hx-trigger="load delay:1s"

If the /messages end point keeps returning a div set up this way, it will keep “polling” back to the URL every second.

Load polling can be useful in situations where a poll has an end point at which point the polling terminates, such as when you are showing the user a progress bar.

#Request Indicators

When an AJAX request is issued it is often good to let the user know that something is happening since the browser will not give them any feedback. You can accomplish this in htmx by using htmx-indicator class.

The htmx-indicator class is defined so that the opacity of any element with this class is 0 by default, making it invisible but present in the DOM.

When htmx issues a request, it will put a htmx-request class onto an element (either the requesting element or another element, if specified). The htmx-request class will cause a child element with the htmx-indicator class on it to transition to an opacity of 1, showing the indicator.

<button hx-get="/click">
    Click Me!
    <img class="htmx-indicator" src="/spinner.gif">

Here we have a button. When it is clicked the htmx-request class will be added to it, which will reveal the spinner gif element. (I like SVG spinners these days.)

While the htmx-indicator class uses opacity to hide and show the progress indicator, if you would prefer another mechanism you can create your own CSS transition like so:

.htmx-request .htmx-indicator{

If you want the htmx-request class added to a different element, you can use the hx-indicator attribute with a CSS selector to do so:

    <button hx-get="/click" hx-indicator="#indicator">
        Click Me!
    <img id="indicator" class="htmx-indicator" src="/spinner.gif"/>

Here we call out the indicator explicitly by id. Note that we could have placed the class on the parent div as well and had the same effect.

You can also add the the disabled attribute to elements for the duration of a request by using the hx-disabled-elt attribute.


If you want the response to be loaded into a different element other than the one that made the request, you can use the hx-target attribute, which takes a CSS selector. Looking back at our Live Search example:

<input type="text" name="q"
    hx-trigger="keyup delay:500ms changed"
<div id="search-results"></div>

You can see that the results from the search are going to be loaded into div#search-results, rather than into the input tag.

#Extended CSS Selectors

hx-target, and most attributes that take a CSS selector, support an “extended” CSS syntax:

  • You can use the this keyword, which indicates that the element that the hx-target attribute is on is the target
  • The closest <CSS selector> syntax will find the closest ancestor element or itself, that matches the given CSS selector. (e.g. closest tr will target the closest table row to the element)
  • The next <CSS selector> syntax will find the next element in the DOM matching the given CSS selector.
  • The previous <CSS selector> syntax will find the previous element in the DOM the given CSS selector.
  • find <CSS selector> which will find the first child descendant element that matches the given CSS selector. (e.g find tr would target the first child descendant row to the element)

In addition, a CSS selector may be wrapped in < and /> characters, mimicking the query literal syntax of hyperscript.

Relative targets like this can be useful for creating flexible user interfaces without peppering your DOM with loads of id attributes.


htmx offers a few different ways to swap the HTML returned into the DOM. By default, the content replaces the innerHTML of the target element. You can modify this by using the hx-swap attribute with any of the following values:

innerHTMLthe default, puts the content inside the target element
outerHTMLreplaces the entire target element with the returned content
afterbeginprepends the content before the first child inside the target
beforebeginprepends the content before the target in the targets parent element
beforeendappends the content after the last child inside the target
afterendappends the content after the target in the targets parent element
deletedeletes the target element regardless of the response
nonedoes not append content from response (Out of Band Swaps and Response Headers will still be processed)

#Morph Swaps

In addition to the standard swap mechanisms above, htmx also supports morphing swaps, via extensions. Morphing swaps attempt to merge new content into the existing DOM, rather than simply replacing it. They often do a better job preserving things like focus, video state, etc. by mutating existing nodes in-place during the swap operation, at the cost of more CPU.

The following extensions are available for morph-style swaps:

#View Transitions

The new, experimental View Transitions API gives developers a way to create an animated transition between different DOM states. It is still in active development and is not available in all browsers, but htmx provides a way to work with this new API that falls back to the non-transition mechanism if the API is not available in a given browser.

You can experiment with this new API using the following approaches:

  • Set the htmx.config.globalViewTransitions config variable to true to use transitions for all swaps
  • Use the transition:true option in the hx-swap attribute
  • If an element swap is going to be transitioned due to either of the above configurations, you may catch the htmx:beforeTransition event and call preventDefault() on it to cancel the transition.

View Transitions can be configured using CSS, as outlined in the Chrome documentation for the feature.

You can see a view transition example on the Animation Examples page.

#Swap Options

The hx-swap attribute supports many options for tuning the swapping behavior of htmx. For example, by default htmx will swap in the title of a title tag found anywhere in the new content. You can turn this behavior off by setting the ignoreTitle modifier to true:

    <button hx-post="/like" hx-swap="outerHTML ignoreTitle:true">Like</button>

The modifiers available on hx-swap are:

transitiontrue or false, whether to use the view transition API for this swap
swapThe swap delay to use (e.g. 100ms) between when old content is cleared and the new content is inserted
settleThe settle delay to use (e.g. 100ms) between when new content is inserted and when it is settled
ignoreTitleIf set to true, any title found in the new content will be ignored and not update the document title
scrolltop or bottom, will scroll the target element to its top or bottom
showtop or bottom, will scroll the target elements top or bottom into view

All swap modifiers appear after the swap style is specified, and are colon-separated.

See the hx-swap documentation for more details on these options.


Often you want to coordinate the requests between two elements. For example, you may want a request from one element to supersede the request of another element, or to wait until the other elements request has finished.

htmx offers a hx-sync attribute to help you accomplish this.

Consider a race condition between a form submission and an individual input’s validation request in this HTML:

<form hx-post="/store">
    <input id="title" name="title" type="text"
    <button type="submit">Submit</button>

Without using hx-sync, filling out the input and immediately submitting the form triggers two parallel requests to /validate and /store.

Using hx-sync="closest form:abort" on the input will watch for requests on the form and abort the input’s request if a form request is present or starts while the input request is in flight:

<form hx-post="/store">
    <input id="title" name="title" type="text"
        hx-sync="closest form:abort"
    <button type="submit">Submit</button>

This resolves the synchronization between the two elements in a declarative way.

htmx also supports a programmatic way to cancel requests: you can send the htmx:abort event to an element to cancel any in-flight requests:

<button id="request-button" hx-post="/example">
    Issue Request
<button onclick="htmx.trigger('#request-button', 'htmx:abort')">
    Cancel Request

More examples and details can be found on the hx-sync attribute page.

#CSS Transitions

htmx makes it easy to use CSS Transitions without javascript. Consider this HTML content:

<div id="div1">Original Content</div>

Imagine this content is replaced by htmx via an ajax request with this new content:

<div id="div1" class="red">New Content</div>

Note two things:

  • The div has the same id in the original and in the new content
  • The red class has been added to the new content

Given this situation, we can write a CSS transition from the old state to the new state:

.red {
    color: red;
    transition: all ease-in 1s ;

When htmx swaps in this new content, it will do so in such a way that the CSS transition will apply to the new content, giving you a nice, smooth transition to the new state.

So, in summary, all you need to do to use CSS transitions for an element is keep its id stable across requests!

You can see the Animation Examples for more details and live demonstrations.


To understand how CSS transitions actually work in htmx, you must understand the underlying swap & settle model that htmx uses.

When new content is received from a server, before the content is swapped in, the existing content of the page is examined for elements that match by the id attribute. If a match is found for an element in the new content, the attributes of the old content are copied onto the new element before the swap occurs. The new content is then swapped in, but with the old attribute values. Finally, the new attribute values are swapped in, after a “settle” delay (20ms by default). A little crazy, but this is what allows CSS transitions to work without any javascript by the developer.

#Out of Band Swaps

If you want to swap content from a response directly into the DOM by using the id attribute you can use the hx-swap-oob attribute in the response html:

<div id="message" hx-swap-oob="true">Swap me directly!</div>
Additional Content

In this response, div#message would be swapped directly into the matching DOM element, while the additional content would be swapped into the target in the normal manner.

You can use this technique to “piggy-back” updates on other requests.

#Selecting Content To Swap

If you want to select a subset of the response HTML to swap into the target, you can use the hx-select attribute, which takes a CSS selector and selects the matching elements from the response.

You can also pick out pieces of content for an out-of-band swap by using the hx-select-oob attribute, which takes a list of element IDs to pick out and swap.

#Preserving Content During A Swap

If there is content that you wish to be preserved across swaps (e.g. a video player that you wish to remain playing even if a swap occurs) you can use the hx-preserve attribute on the elements you wish to be preserved.


By default, an element that causes a request will include its value if it has one. If the element is a form it will include the values of all inputs within it.

As with HTML forms, the name attribute of the input is used as the parameter name in the request that htmx sends.

Additionally, if the element causes a non-GET request, the values of all the inputs of the nearest enclosing form will be included.

If you wish to include the values of other elements, you can use the hx-include attribute with a CSS selector of all the elements whose values you want to include in the request.

If you wish to filter out some parameters you can use the hx-params attribute.

Finally, if you want to programmatically modify the parameters, you can use the htmx:configRequest event.

#File Upload

If you wish to upload files via an htmx request, you can set the hx-encoding attribute to multipart/form-data. This will use a FormData object to submit the request, which will properly include the file in the request.

Note that depending on your server-side technology, you may have to handle requests with this type of body content very differently.

Note that htmx fires a htmx:xhr:progress event periodically based on the standard progress event during upload, which you can hook into to show the progress of the upload.

See the examples section for more advanced form patterns, including progress bars and error handling.

#Extra Values

You can include extra values in a request using the hx-vals (name-expression pairs in JSON format) and hx-vars attributes (comma-separated name-expression pairs that are dynamically computed).

#Confirming Requests

Often you will want to confirm an action before issuing a request. htmx supports the hx-confirm attribute, which allows you to confirm an action using a simple javascript dialog:

<button hx-delete="/account" hx-confirm="Are you sure you wish to delete your account?">
    Delete My Account

Using events you can implement more sophisticated confirmation dialogs. The confirm example shows how to use sweetalert2 library for confirmation of htmx actions.

#Attribute Inheritance

Most attributes in htmx are inherited: they apply to the element they are on as well as any children elements. This allows you to “hoist” attributes up the DOM to avoid code duplication. Consider the following htmx:

<button hx-delete="/account" hx-confirm="Are you sure?">
    Delete My Account
<button hx-put="/account" hx-confirm="Are you sure?">
    Update My Account

Here we have a duplicate hx-confirm attribute. We can hoist this attribute to a parent element:

<div hx-confirm="Are you sure?">
    <button hx-delete="/account">
        Delete My Account
    <button hx-put="/account">
        Update My Account

This hx-confirm attribute will now apply to all htmx-powered elements within it.

Sometimes you wish to undo this inheritance. Consider if we had a cancel button to this group, but didn’t want it to be confirmed. We could add an unset directive on it like so:

<div hx-confirm="Are you sure?">
    <button hx-delete="/account">
        Delete My Account
    <button hx-put="/account">
        Update My Account
    <button hx-confirm="unset" hx-get="/">

The top two buttons would then show a confirm dialog, but the bottom cancel button would not.

Automatic inheritance can be disabled using the hx-disinherit attribute.


Htmx supports “boosting” regular HTML anchors and forms with the hx-boost attribute. This attribute will convert all anchor tags and forms into AJAX requests that, by default, target the body of the page.

Here is an example:

<div hx-boost="true">
    <a href="/blog">Blog</a>

The anchor tag in this div will issue an AJAX GET request to /blog and swap the response into the body tag.

#Progressive Enhancement

A feature of hx-boost is that it degrades gracefully if javascript is not enabled: the links and forms continue to work, they simply don’t use ajax requests. This is known as Progressive Enhancement, and it allows a wider audience to use your sites functionality.

Other htmx patterns can be adapted to achieve progressive enhancement as well, but they will require more thought.

Consider the active search example. As it is written, it will not degrade gracefully: someone who does not have javascript enabled will not be able to use this feature. This is done for simplicity’s sake, to keep the example as brief as possible.

However, you could wrap the htmx-enhanced input in a form element:

<form action="/search" method="POST">
    <input class="form-control" type="search"
        name="search" placeholder="Begin typing to search users..."
        hx-trigger="keyup changed delay:500ms, search"

With this in place, javascript-enabled clients would still get the nice active-search UX, but non-javascript enabled clients would be able to hit the enter key and still search. Even better, you could add a “Search” button as well. You would then need to update the form with an hx-post that mirrored the action attribute, or perhaps use hx-boost on it.

You would need to check on the server side for the HX-Request header to differentiate between an htmx-driven and a regular request, to determine exactly what to render to the client.

Other patterns can be adapted similarly to achieve the progressive enhancement needs of your application.

As you can see, this requires more thought and more work. It also rules some functionality entirely out of bounds. These tradeoffs must be made by you, the developer, with respect to your projects goals and audience.

Accessibility is a concept closely related to progressive enhancement. Using progressive enhancement techniques such as hx-boost will make your htmx application more accessible to a wide array of users.

htmx-based applications are very similar to normal, non-AJAX driven web applications because htmx is HTML-oriented.

As such, the normal HTML accessibility recommendations apply. For example:

  • Use semantic HTML as much as possible (i.e. the right tags for the right things)
  • Ensure focus state is clearly visible
  • Associate text labels with all form fields
  • Maximize the readability of your application with appropriate fonts, contrast, etc.

#Web Sockets & SSE

htmx has experimental support for declarative use of both WebSockets and Server Sent Events.

Note: In htmx 2.0, these features will be migrated to extensions. These new extensions are already available in htmx 1.7+ and, if you are writing new code, you are encouraged to use the extensions instead. All new feature work for both SSE and web sockets will be done in the extensions.

Please visit the SSE extension and WebSocket extension pages to learn more about the new extensions.


If you wish to establish a WebSocket connection in htmx, you use the hx-ws attribute:

<div hx-ws="connect:wss:/chatroom">
    <div id="chat_room">
    <form hx-ws="send:submit">
        <input name="chat_message">

The connect declaration established the connection, and the send declaration tells the form to submit values to the socket on submit.

More details can be found on the hx-ws attribute page

#Server Sent Events

Server Sent Events are a way for servers to send events to browsers. It provides a higher-level mechanism for communication between the server and the browser than websockets.

If you want an element to respond to a Server Sent Event via htmx, you need to do two things:

  1. Define an SSE source. To do this, add a hx-sse attribute on a parent element with a connect:<url> declaration that specifies the URL from which Server Sent Events will be received.

  2. Define elements that are descendents of this element that are triggered by server sent events using the hx-trigger="sse:<event_name>" syntax

Here is an example:

<body hx-sse="connect:/news_updates">
    <div hx-trigger="sse:new_news" hx-get="/news"></div>

Depending on your implementation, this may be more efficient than the polling example above since the server would notify the div if there was new news to get, rather than the steady requests that a poll causes.

#History Support

Htmx provides a simple mechanism for interacting with the browser history API:

If you want a given element to push its request URL into the browser navigation bar and add the current state of the page to the browser’s history, include the hx-push-url attribute:

<a hx-get="/blog" hx-push-url="true">Blog</a>

When a user clicks on this link, htmx will snapshot the current DOM and store it before it makes a request to /blog. It then does the swap and pushes a new location onto the history stack.

When a user hits the back button, htmx will retrieve the old content from storage and swap it back into the target, simulating “going back” to the previous state. If the location is not found in the cache, htmx will make an ajax request to the given URL, with the header HX-History-Restore-Request set to true, and expects back the HTML needed for the entire page. Alternatively, if the htmx.config.refreshOnHistoryMiss config variable is set to true, it will issue a hard browser refresh.

NOTE: If you push a URL into the history, you must be able to navigate to that URL and get a full page back! A user could copy and paste the URL into an email, or new tab. Additionally, htmx will need the entire page when restoring history if the page is not in the history cache.

#Specifying History Snapshot Element

By default, htmx will use the body to take and restore the history snapshot from. This is usually the right thing, but if you want to use a narrower element for snapshotting you can use the hx-history-elt attribute to specify a different one.

Careful: this element will need to be on all pages or restoring from history won’t work reliably.

#Undoing DOM Mutations By 3rd Party Libraries

If you are using a 3rd party library and want to use the htmx history feature, you will need to clean up the DOM before a snapshot is taken. Let’s consider the Tom Select library, which makes select elements a much richer user experience. Let’s set up TomSelect to turn any input element with the .tomselect class into a rich select element.

First we need to initialize elements that have the class in new content:

htmx.onLoad(function (target) {
    // find all elements in the new content that should be
    // an editor and init w/ quill
    var editors = target.querySelectorAll(".tomselect")
            .forEach(elt => new TomSelect(elt))

This will create a rich selector for all input elements that have the .tomselect class on it. However, it mutates the DOM and we don’t want that mutation saved to the history cache, since TomSelect will be reinitialized when the history content is loaded back into the screen.

To deal with this, we need to catch the htmx:beforeHistorySave event and clean out the TomSelect mutations by calling destroy() on them:

htmx.on('htmx:beforeHistorySave', function() {
    // find all TomSelect elements
            .forEach(elt => elt.tomselect.destroy()) // and call destroy() on them

This will revert the DOM to the original HTML, thus allowing for a clean snapshot.

#Disabling History Snapshots

History snapshotting can be disabled for a URL by setting the hx-history attribute to false on any element in the current document, or any html fragment loaded into the current document by htmx. This can be used to prevent sensitive data entering the localStorage cache, which can be important for shared-use / public computers. History navigation will work as expected, but on restoration the URL will be requested from the server instead of the local history cache.

#Requests & Responses

Htmx expects responses to the AJAX requests it makes to be HTML, typically HTML fragments (although a full HTML document, matched with a hx-select tag can be useful too). Htmx will then swap the returned HTML into the document at the target specified and with the swap strategy specified.

Sometimes you might want to do nothing in the swap, but still perhaps trigger a client side event (see below). For this situation you can return a 204 - No Content response code, and htmx will ignore the content of the response.

In the event of an error response from the server (e.g. a 404 or a 501), htmx will trigger the htmx:responseError event, which you can handle.

In the event of a connection error, the htmx:sendError event will be triggered.


When using htmx in a cross origin context, remember to configure your web server to set Access-Control headers in order for htmx headers to be visible on the client side.

See all the request and response headers that htmx implements.

#Request Headers

htmx includes a number of useful headers in requests:

HX-Boostedindicates that the request is via an element using hx-boost
HX-Current-URLthe current URL of the browser
HX-History-Restore-Request“true” if the request is for history restoration after a miss in the local history cache
HX-Promptthe user response to an hx-prompt
HX-Requestalways “true”
HX-Targetthe id of the target element if it exists
HX-Trigger-Namethe name of the triggered element if it exists
HX-Triggerthe id of the triggered element if it exists

#Response Headers

htmx supports some htmx-specific response headers:

  • HX-Location - allows you to do a client-side redirect that does not do a full page reload
  • HX-Push-Url - pushes a new url into the history stack
  • HX-Redirect - can be used to do a client-side redirect to a new location
  • HX-Refresh - if set to “true” the client-side will do a full refresh of the page
  • HX-Replace-Url - replaces the current URL in the location bar
  • HX-Reswap - allows you to specify how the response will be swapped. See hx-swap for possible values
  • HX-Retarget - a CSS selector that updates the target of the content update to a different element on the page
  • HX-Reselect - a CSS selector that allows you to choose which part of the response is used to be swapped in. Overrides an existing hx-select on the triggering element
  • HX-Trigger - allows you to trigger client-side events
  • HX-Trigger-After-Settle - allows you to trigger client-side events after the settle step
  • HX-Trigger-After-Swap - allows you to trigger client-side events after the swap step

For more on the HX-Trigger headers, see HX-Trigger Response Headers.

Submitting a form via htmx has the benefit of no longer needing the Post/Redirect/Get Pattern. After successfully processing a POST request on the server, you don’t need to return a HTTP 302 (Redirect). You can directly return the new HTML fragment.

#Request Order of Operations

The order of operations in a htmx request are:

  • The element is triggered and begins a request
    • Values are gathered for the request
    • The htmx-request class is applied to the appropriate elements
    • The request is then issued asynchronously via AJAX
      • Upon getting a response the target element is marked with the htmx-swapping class
      • An optional swap delay is applied (see the hx-swap attribute)
      • The actual content swap is done
        • the htmx-swapping class is removed from the target
        • the htmx-added class is added to each new piece of content
        • the htmx-settling class is applied to the target
        • A settle delay is done (default: 20ms)
        • The DOM is settled
        • the htmx-settling class is removed from the target
        • the htmx-added class is removed from each new piece of content

You can use the htmx-swapping and htmx-settling classes to create CSS transitions between pages.


Htmx integrates with the HTML5 Validation API and will not issue a request for a form if a validatable input is invalid. This is true for both AJAX requests as well as WebSocket sends.

Htmx fires events around validation that can be used to hook in custom validation and error handling:

  • htmx:validation:validate - called before an elements checkValidity() method is called. May be used to add in custom validation logic
  • htmx:validation:failed - called when checkValidity() returns false, indicating an invalid input
  • htmx:validation:halted - called when a request is not issued due to validation errors. Specific errors may be found in the event.detail.errors object

Non-form elements do not validate before they make requests by default, but you can enable validation by setting the hx-validate attribute to “true”.

#Validation Example

Here is an example of an input that uses the hx-on attribute to catch the htmx:validation:validate event and require that the input have the value foo:

<form id="example-form" hx-post="/test">
    <input name="example"
           onkeyup="this.setCustomValidity('') // reset the validation on keyup"
           hx-on:htmx:validation:validate="if(this.value != 'foo') {
                    this.setCustomValidity('Please enter the value foo') // set the validation error
                    htmx.find('#foo-form').reportValidity()              // report the issue

Note that all client side validations must be re-done on the server side, as they can always be bypassed.


Htmx allows you to use CSS transitions in many situations using only HTML and CSS.

Please see the Animation Guide for more details on the options available.


Htmx has an extension mechanism that allows you to customize the libraries’ behavior. Extensions are defined in javascript and then used via the hx-ext attribute:

<div hx-ext="debug">
    <button hx-post="/example">This button used the debug extension</button>
    <button hx-post="/example" hx-ext="ignore:debug">This button does not</button>

If you are interested in adding your own extension to htmx, please see the extension docs

#Included Extensions

Htmx includes some extensions that are tested against the htmx code base. Here are a few:

json-encuse JSON encoding in the body of requests, rather than the default x-www-form-urlencoded
morphdom-swapan extension for using the morphdom library as the swapping mechanism in htmx.
alpine-morphan extension for using the Alpine.js morph plugin as the swapping mechanism in htmx.
client-side-templatessupport for client side template processing of JSON responses
path-depsan extension for expressing path-based dependencies similar to intercoolerjs
class-toolsan extension for manipulating timed addition and removal of classes on HTML elements
multi-swapallows to swap multiple elements with different swap methods
response-targetsallows to swap elements for responses with HTTP codes beyond 200

See the extensions page for a complete list.

#Events & Logging

Htmx has an extensive events mechanism, which doubles as the logging system.

If you want to register for a given htmx event you can use

document.body.addEventListener('htmx:load', function(evt) {

or, if you would prefer, you can use the following htmx helper:

htmx.on("htmx:load", function(evt) {

The htmx:load event is fired every time an element is loaded into the DOM by htmx, and is effectively the equivalent to the normal load event.

Some common uses for htmx events are:

#Initialize A 3rd Party Library With Events

Using the htmx:load event to initialize content is so common that htmx provides a helper function:

htmx.onLoad(function(target) {

This does the same thing as the first example, but is a little cleaner.

#Configure a Request With Events

You can handle the htmx:configRequest event in order to modify an AJAX request before it is issued:

document.body.addEventListener('htmx:configRequest', function(evt) {
    evt.detail.parameters['auth_token'] = getAuthToken(); // add a new parameter into the request
    evt.detail.headers['Authentication-Token'] = getAuthToken(); // add a new header into the request

Here we add a parameter and header to the request before it is sent.

#Modifying Swapping Behavior With Events

You can handle the htmx:beforeSwap event in order to modify the swap behavior of htmx:

document.body.addEventListener('htmx:beforeSwap', function(evt) {
    if(evt.detail.xhr.status === 404){
        // alert the user when a 404 occurs (maybe use a nicer mechanism than alert())
        alert("Error: Could Not Find Resource");
    } else if(evt.detail.xhr.status === 422){
        // allow 422 responses to swap as we are using this as a signal that
        // a form was submitted with bad data and want to rerender with the
        // errors
        // set isError to false to avoid error logging in console
        evt.detail.shouldSwap = true;
        evt.detail.isError = false;
    } else if(evt.detail.xhr.status === 418){
        // if the response code 418 (I'm a teapot) is returned, retarget the
        // content of the response to the element with the id `teapot`
        evt.detail.shouldSwap = true; = htmx.find("#teapot");

Here we handle a few 400-level error response codes that would normally not do a swap in htmx.

#Event Naming

Note that all events are fired with two different names

  • Camel Case
  • Kebab Case

So, for example, you can listen for htmx:afterSwap or for htmx:after-swap. This facilitates interoperability with other libraries. Alpine.js, for example, requires kebab case.


If you set a logger at htmx.logger, every event will be logged. This can be very useful for troubleshooting:

htmx.logger = function(elt, event, data) {
    if(console) {
        console.log(event, elt, data);


Declarative and event driven programming with htmx (or any other declarative language) can be a wonderful and highly productive activity, but one disadvantage when compared with imperative approaches is that it can be trickier to debug.

Figuring out why something isn’t happening, for example, can be difficult if you don’t know the tricks.

Well, here are the tricks:

The first debugging tool you can use is the htmx.logAll() method. This will log every event that htmx triggers and will allow you to see exactly what the library is doing.


Of course, that won’t tell you why htmx isn’t doing something. You might also not know what events a DOM element is firing to use as a trigger. To address this, you can use the monitorEvents() method available in the browser console:


This will spit out all events that are occurring on the element with the id theElement to the console, and allow you to see exactly what is going on with it.

Note that this only works from the console, you cannot embed it in a script tag on your page.

Finally, push come shove, you might want to just debug htmx.js by loading up the unminimized version. It’s about 2500 lines of javascript, so not an insurmountable amount of code. You would most likely want to set a break point in the issueAjaxRequest() and handleAjaxResponse() methods to see what’s going on.

And always feel free to jump on the Discord if you need help.

#Creating Demos

Sometimes, in order to demonstrate a bug or clarify a usage, it is nice to be able to use a javascript snippet site like jsfiddle. To facilitate easy demo creation, htmx hosts a demo script site that will install:

  • htmx
  • hyperscript
  • a request mocking library

Simply add the following script tag to your demo/fiddle/whatever:

<script src=""></script>

This helper allows you to add mock responses by adding template tags with a url attribute to indicate which URL. The response for that url will be the innerHTML of the template, making it easy to construct mock responses. You can add a delay to the response with a delay attribute, which should be an integer indicating the number of milliseconds to delay

You may embed simple expressions in the template with the ${} syntax.

Note that this should only be used for demos and is in no way guaranteed to work for long periods of time as it will always be grabbing the latest versions htmx and hyperscript!

#Demo Example

Here is an example of the code in action:

<!-- load demo environment -->
<script src=""></script>

<!-- post to /foo -->
<button hx-post="/foo" hx-target="#result">
    Count Up
<output id="result"></output>

<!-- respond to /foo with some dynamic content in a template tag -->
    globalInt = 0;
<template url="/foo" delay="500"> <!-- note the url and delay attributes -->


While htmx encourages a hypermedia approach to building web applications, it offers many options for client scripting. Scripting is included in the REST-ful description of web architecture, see: Code-On-Demand. As much as is feasible, we recommend a hypermedia-friendly approach to scripting in your web application:

The primary integration point between htmx and scripting solutions is the events that htmx sends and can respond to. See the SortableJS example in the 3rd Party Javascript section for a good template for integrating a JavaScript library with htmx via events.

Scripting solutions that pair well with htmx include:

  • VanillaJS - Simply using the built-in abilities of JavaScript to hook in event handlers to respond to the events htmx emits can work very well for scripting. This is an extremely lightweight and increasingly popular approach.
  • AlpineJS - Alpine.js provides a rich set of tools for creating sophisticated front end scripts, including reactive programming support, while still remaining extremely lightweight. Alpine encourages the “inline scripting” approach that we feel pairs well with htmx.
  • jQuery - Despite its age and reputation in some circles, jQuery pairs very well with htmx, particularly in older code-bases that already have a lot of jQuery in them.
  • hyperscript - Hyperscript is an experimental front-end scripting language created by the same team that created htmx. It is designed to embed well in HTML and both respond to and create events, and pairs very well with htmx.

We have an entire chapter entitled “Client-Side Scripting” in our book that looks at how scripting can be integrated into your htmx-based application.

#The hx-on* Attributes

HTML allows the embedding of inline scripts via the onevent properties, such as onClick:

<button onclick="alert('You clicked me!')">
    Click Me!

This feature allows scripting logic to be co-located with the HTML elements the logic applies to, giving good Locality of Behaviour (LoB). Unfortunately, HTML only allows on* attributes for a fixed number of specific DOM events (e.g. onclick) and doesn’t provide a generalized mechanism for responding to arbitrary events on elements.

In order to address this shortcoming, htmx offers hx-on* attributes. These attributes allow you to respond to any event in a manner that preserves the LoB of the standard on* properties.

If we wanted to respond to the click event using an hx-on attribute, we would write this:

<button hx-on:click="alert('You clicked me!')">
    Click Me!

So, the string hx-on, followed by a colon (or a dash), then by the name of the event.

For a click event, of course, we would recommend sticking with the standard onclick attribute. However, consider an htmx-powered button that wishes to add a parameter to a request using the htmx:config-request event. This would not be possible using a standard on* property, but it can be done using the hx-on:htmx:config-request attribute:

<button hx-post="/example"
        hx-on:htmx:config-request="event.detail.parameters.example = 'Hello Scripting!'">
    Post Me!

Here the example parameter is added to the POST request before it is issued, with the value ‘Hello Scripting!’.

The hx-on* attributes are a very simple mechanism for generalized embedded scripting. It is not a replacement for more fully developed front-end scripting solutions such as AlpineJS or hyperscript. It can, however, augment a VanillaJS-based approach to scripting in your htmx-powered application.

Note that HTML attributes are case insensitive. This means that, unfortunately, events that rely on capitalization/ camel casing, cannot be responded to. If you need to support camel case events we recommend using a more fully functional scripting solution such as AlpineJS or hyperscript. htmx dispatches all its events in both camelCase and in kebab-case for this very reason.

#3rd Party Javascript

Htmx integrates fairly well with third party libraries. If the library fires events on the DOM, you can use those events to trigger requests from htmx.

A good example of this is the SortableJS demo:

<form class="sortable" hx-post="/items" hx-trigger="end">
    <div class="htmx-indicator">Updating...</div>
    <div><input type='hidden' name='item' value='1'/>Item 1</div>
    <div><input type='hidden' name='item' value='2'/>Item 2</div>
    <div><input type='hidden' name='item' value='2'/>Item 3</div>

With Sortable, as with most javascript libraries, you need to initialize content at some point.

In jquery you might do this like so:

$(document).ready(function() {
    var sortables = document.body.querySelectorAll(".sortable");
    for (var i = 0; i < sortables.length; i++) {
        var sortable = sortables[i];
        new Sortable(sortable, {
            animation: 150,
            ghostClass: 'blue-background-class'

In htmx, you would instead use the htmx.onLoad function, and you would select only from the newly loaded content, rather than the entire document:

htmx.onLoad(function(content) {
    var sortables = content.querySelectorAll(".sortable");
    for (var i = 0; i < sortables.length; i++) {
        var sortable = sortables[i];
        new Sortable(sortable, {
            animation: 150,
            ghostClass: 'blue-background-class'

This will ensure that as new content is added to the DOM by htmx, sortable elements are properly initialized.

If javascript adds content to the DOM that has htmx attributes on it, you need to make sure that this content is initialized with the htmx.process() function.

For example, if you were to fetch some data and put it into a div using the fetch API, and that HTML had htmx attributes in it, you would need to add a call to htmx.process() like this:

let myDiv = document.getElementById('my-div')
    .then(response => response.text())
    .then(data => { myDiv.innerHTML = data; htmx.process(myDiv); } );

Some 3rd party libraries create content from HTML template elements. For instance, Alpine JS uses the x-if attribute on templates to add content conditionally. Such templates are not initially part of the DOM and, if they contain htmx attributes, will need a call to htmx.process() after they are loaded. The following example uses Alpine’s $watch function to look for a change of value that would trigger conditional content:

<div x-data="{show_new: false}"
    x-init="$watch('show_new', value => {
        if (show_new) {
    <button @click = "show_new = !show_new">Toggle New Content</button>
    <template x-if="show_new">
        <div id="new_content">
            <a hx-get="/server/newstuff" href="#">New Clickable</a>


htmx works with standard HTTP caching mechanisms out of the box.

If your server adds the Last-Modified HTTP response header to the response for a given URL, the browser will automatically add the If-Modified-Since request HTTP header to the next requests to the same URL. Be mindful that if your server can render different content for the same URL depending on some other headers, you need to use the Vary response HTTP header. For example, if your server renders the full HTML when the HX-Request header is missing or false, and it renders a fragment of that HTML when HX-Request: true, you need to add Vary: HX-Request. That causes the cache to be keyed based on a composite of the response URL and the HX-Request request header — rather than being based just on the response URL.

If you are unable (or unwilling) to use the Vary header, you can alternatively set the configuration parameter getCacheBusterParam to true. If this configuration variable is set, htmx will include a cache-busting parameter in GET requests that it makes, which will prevent browsers from caching htmx-based and non-htmx based responses in the same cache slot.

htmx also works with ETag as expected. Be mindful that if your server can render different content for the same URL (for example, depending on the value of the HX-Request header), the server needs to generate a different ETag for each content.


htmx allows you to define logic directly in your DOM. This has a number of advantages, the largest being Locality of Behavior, which makes your system easier to understand and maintain.

A concern with this approach, however, is security: since htmx increases the expressiveness of HTML, if a malicious user is able to inject HTML into your application, they can leverage this expressiveness of htmx to malicious ends.

#Rule 1: Escape All User Content

The first rule of HTML-based web development has always been: do not trust input from the user. You should escape all 3rd party, untrusted content that is injected into your site. This is to prevent, among other issues, XSS attacks.

There is extensive documentation on XSS and how to prevent it on the excellent OWASP Website, including a Cross Site Scripting Prevention Cheat Sheet.

The good news is that this is a very old and well understood topic, and the vast majority of server-side templating languages support automatic escaping of content to prevent just such an issue.

That being said, there are times people choose to inject HTML more dangerously, often via some sort of raw() mechanism in their templating language. This can be done for good reasons, but if the content being injected is coming from a 3rd party then it must be scrubbed, including removing attributes starting with hx- and data-hx, as well as inline <script> tags, etc.

If you are injecting raw HTML and doing your own escaping, a best practice is to whitelist the attributes and tags you allow, rather than to blacklist the ones you disallow.

#htmx Security Tools

Of course, bugs happen and developers are not perfect, so it is good to have a layered approach to security for your web application, and htmx provides tools to help secure your application as well.

Let’s take a look at them.


The first tool htmx provides to help further secure your application is the hx-disable attribute. This attribute will prevent processing of all htmx attributes on a given element, and on all elements within it. So, for example, if you were including raw HTML content in a template (again, this is not recommended!) then you could place a div around the content with the hx-disable attribute on it:

<div hx-disable>
    <%= raw(user_content) %>

And htmx will not process any htmx-related attributes or features found in that content. This attribute cannot be disabled by injecting further content: if an hx-disable attribute is found anywhere in the parent hierarchy of an element, it will not be processed by htmx.


Another security consideration is htmx history cache. You may have pages that have sensitive data that you do not want stored in the users localStorage cache. You can omit a given page from the history cache by including the hx-history attribute anywhere on the page, and setting its value to false.

#Configuration Options

htmx also provides configuration options related to security:

  • htmx.config.selfRequestsOnly - if set to true, only requests to the same domain as the current document will be allowed
  • htmx.config.allowScriptTags - htmx will process <script> tags found in new content it loads. If you wish to disable this behavior you can set this configuration variable to false
  • htmx.config.historyCacheSize - can be set to 0 to avoid storing any HTML in the localStorage cache
  • htmx.config.allowEval - can be set to false to disable all features of htmx that rely on eval:
    • event filters
    • hx-on: attributes
    • hx-vals with the js: prefix
    • hx-headers with the js: prefix

Note that all features removed by disabling eval() can be reimplemented using your own custom javascript and the htmx event model.


If you want to allow requests to some domains beyond the current host, but not leave things totally open, you can use the htmx:validateUrl event. This event will have the request URL available in the detail.url slot, as well as a sameHost property.

You can inspect these values and, if the request is not valid, invoke preventDefault() on the event to prevent the request from being issued.

document.body.addEventListener('htmx:validateUrl', function (evt) {
  // only allow requests to the current server as well as
  if (!evt.detail.sameHost && evt.detail.url.hostname !== "") {

#CSP Options

Browsers also provide tools for further securing your web application. The most powerful tool available is a Content Security Policy. Using a CSP you can tell the browser to, for example, not issue requests to non-origin hosts, to not evaluate inline script tags, etc.

Here is an example CSP in a meta tag:

    <meta http-equiv="Content-Security-Policy" content="default-src 'self';">

This tells the browser “Only allow connections to the original (source) domain”. This would be redundant with the htmx.config.selfRequestsOnly, but a layered approach to security is warranted and, in fact, ideal, when dealing with application security.

A full discussion of CSPs is beyond the scope of this document, but the MDN Article provide a good jumping off point for exploring this topic.

#Configuring htmx

Htmx has some configuration options that can be accessed either programmatically or declaratively. They are listed below:

Config VariableInfo
htmx.config.historyEnableddefaults to true, really only useful for testing
htmx.config.historyCacheSizedefaults to 10
htmx.config.refreshOnHistoryMissdefaults to false, if set to true htmx will issue a full page refresh on history misses rather than use an AJAX request
htmx.config.defaultSwapStyledefaults to innerHTML
htmx.config.defaultSwapDelaydefaults to 0
htmx.config.defaultSettleDelaydefaults to 20
htmx.config.includeIndicatorStylesdefaults to true (determines if the indicator styles are loaded)
htmx.config.indicatorClassdefaults to htmx-indicator
htmx.config.requestClassdefaults to htmx-request
htmx.config.addedClassdefaults to htmx-added
htmx.config.settlingClassdefaults to htmx-settling
htmx.config.swappingClassdefaults to htmx-swapping
htmx.config.allowEvaldefaults to true, can be used to disable htmx’s use of eval for certain features (e.g. trigger filters)
htmx.config.allowScriptTagsdefaults to true, determines if htmx will process script tags found in new content
htmx.config.inlineScriptNoncedefaults to '', meaning that no nonce will be added to inline scripts
htmx.config.attributesToSettledefaults to ["class", "style", "width", "height"], the attributes to settle during the settling phase
htmx.config.useTemplateFragmentsdefaults to false, HTML template tags for parsing content from the server (not IE11 compatible!)
htmx.config.wsReconnectDelaydefaults to full-jitter
htmx.config.wsBinaryTypedefaults to blob, the the type of binary data being received over the WebSocket connection
htmx.config.disableSelectordefaults to [hx-disable], [data-hx-disable], htmx will not process elements with this attribute on it or a parent
htmx.config.withCredentialsdefaults to false, allow cross-site Access-Control requests using credentials such as cookies, authorization headers or TLS client certificates
htmx.config.timeoutdefaults to 0, the number of milliseconds a request can take before automatically being terminated
htmx.config.scrollBehaviordefaults to ‘smooth’, the behavior for a boosted link on page transitions. The allowed values are auto and smooth. Smooth will smoothscroll to the top of the page while auto will behave like a vanilla link.
htmx.config.defaultFocusScrollif the focused element should be scrolled into view, defaults to false and can be overridden using the focus-scroll swap modifier.
htmx.config.getCacheBusterParamdefaults to false, if set to true htmx will include a cache-busting parameter in GET requests to avoid caching partial responses by the browser
htmx.config.globalViewTransitionsif set to true, htmx will use the View Transition API when swapping in new content.
htmx.config.methodsThatUseUrlParamsdefaults to ["get"], htmx will format requests with these methods by encoding their parameters in the URL, not the request body
htmx.config.selfRequestsOnlydefaults to false, if set to true will only allow AJAX requests to the same domain as the current document
htmx.config.ignoreTitledefaults to false, if set to true htmx will not update the title of the document when a title tag is found in new content
htmx.config.scrollIntoViewOnBoostdefaults to true, whether or not the target of a boosted element is scrolled into the viewport. If hx-target is omitted on a boosted element, the target defaults to body, causing the page to scroll to the top.
htmx.config.triggerSpecsCachedefaults to null, the cache to store evaluated trigger specifications into, improving parsing performance at the cost of more memory usage. You may define a simple object to use a never-clearing cache, or implement your own system using a proxy object

You can set them directly in javascript, or you can use a meta tag:

<meta name="htmx-config" content='{"defaultSwapStyle":"outerHTML"}'>


And that’s it!

Have fun with htmx! You can accomplish quite a bit without writing a lot of code!