Promise.all()

Baseline Widely available

This feature is well established and works across many devices and browser versions. It’s been available across browsers since July 2015.

The Promise.all() static method takes an iterable of promises as input and returns a single Promise. This returned promise fulfills when all of the input's promises fulfill (including when an empty iterable is passed), with an array of the fulfillment values. It rejects when any of the input's promises rejects, with this first rejection reason.

Try it

Syntax

js
Promise.all(iterable)

Parameters

iterable

An iterable (such as an Array) of promises.

Return value

A Promise that is:

  • Already fulfilled, if the iterable passed is empty.
  • Asynchronously fulfilled, when all the promises in the given iterable fulfill. The fulfillment value is an array of fulfillment values, in the order of the promises passed, regardless of completion order. If the iterable passed is non-empty but contains no pending promises, the returned promise is still asynchronously (instead of synchronously) fulfilled.
  • Asynchronously rejected, when any of the promises in the given iterable rejects. The rejection reason is the rejection reason of the first promise that was rejected.

Description

The Promise.all() method is one of the promise concurrency methods. It can be useful for aggregating the results of multiple promises. It is typically used when there are multiple related asynchronous tasks that the overall code relies on to work successfully — all of whom we want to fulfill before the code execution continues.

Promise.all() will reject immediately upon any of the input promises rejecting. In comparison, the promise returned by Promise.allSettled() will wait for all input promises to complete, regardless of whether or not one rejects. Use allSettled() if you need the final result of every promise in the input iterable.

Examples

Using Promise.all()

Promise.all waits for all fulfillments (or the first rejection).

js
const p1 = Promise.resolve(3);
const p2 = 1337;
const p3 = new Promise((resolve, reject) => {
  setTimeout(() => {
    resolve("foo");
  }, 100);
});

Promise.all([p1, p2, p3]).then((values) => {
  console.log(values); // [3, 1337, "foo"]
});

If the iterable contains non-promise values, they will be ignored, but still counted in the returned promise array value (if the promise is fulfilled):

js
// All values are non-promises, so the returned promise gets fulfilled
const p = Promise.all([1, 2, 3]);
// The only input promise is already fulfilled,
// so the returned promise gets fulfilled
const p2 = Promise.all([1, 2, 3, Promise.resolve(444)]);
// One (and the only) input promise is rejected,
// so the returned promise gets rejected
const p3 = Promise.all([1, 2, 3, Promise.reject(555)]);

// Using setTimeout, we can execute code after the queue is empty
setTimeout(() => {
  console.log(p);
  console.log(p2);
  console.log(p3);
});

// Logs:
// Promise { <state>: "fulfilled", <value>: Array[3] }
// Promise { <state>: "fulfilled", <value>: Array[4] }
// Promise { <state>: "rejected", <reason>: 555 }

Asynchronicity or synchronicity of Promise.all

This following example demonstrates the asynchronicity of Promise.all when a non-empty iterable is passed:

js
// Passing an array of promises that are already resolved,
// to trigger Promise.all as soon as possible
const resolvedPromisesArray = [Promise.resolve(33), Promise.resolve(44)];

const p = Promise.all(resolvedPromisesArray);
// Immediately logging the value of p
console.log(p);

// Using setTimeout, we can execute code after the queue is empty
setTimeout(() => {
  console.log("the queue is now empty");
  console.log(p);
});

// Logs, in order:
// Promise { <state>: "pending" }
// the queue is now empty
// Promise { <state>: "fulfilled", <value>: Array[2] }

The same thing happens if Promise.all rejects:

js
const mixedPromisesArray = [Promise.resolve(33), Promise.reject(44)];
const p = Promise.all(mixedPromisesArray);
console.log(p);
setTimeout(() => {
  console.log("the queue is now empty");
  console.log(p);
});

// Logs:
// Promise { <state>: "pending" }
// the queue is now empty
// Promise { <state>: "rejected", <reason>: 44 }

Promise.all resolves synchronously if and only if the iterable passed is empty:

js
const p = Promise.all([]); // Will be immediately resolved
const p2 = Promise.all([1337, "hi"]); // Non-promise values are ignored, but the evaluation is done asynchronously
console.log(p);
console.log(p2);
setTimeout(() => {
  console.log("the queue is now empty");
  console.log(p2);
});

// Logs:
// Promise { <state>: "fulfilled", <value>: Array[0] }
// Promise { <state>: "pending" }
// the queue is now empty
// Promise { <state>: "fulfilled", <value>: Array[2] }

Using Promise.all() with async functions

Within async functions, it's very common to "over-await" your code. For example, given the following functions:

js
function promptForDishChoice() {
  return new Promise((resolve, reject) => {
    const dialog = document.createElement("dialog");
    dialog.innerHTML = `
<form method="dialog">
  <p>What would you like to eat?</p>
  <select>
    <option value="pizza">Pizza</option>
    <option value="pasta">Pasta</option>
    <option value="salad">Salad</option>
  </select>
  <menu>
    <li><button value="cancel">Cancel</button></li>
    <li><button type="submit" value="ok">OK</button></li>
  </menu>
</form>
    `;
    dialog.addEventListener("close", () => {
      if (dialog.returnValue === "ok") {
        resolve(dialog.querySelector("select").value);
      } else {
        reject(new Error("User cancelled dialog"));
      }
    });
    document.body.appendChild(dialog);
    dialog.showModal();
  });
}

async function fetchPrices() {
  const response = await fetch("/prices");
  return await response.json();
}

You may write a function like this:

js
async function getPrice() {
  const choice = await promptForDishChoice();
  const prices = await fetchPrices();
  return prices[choice];
}

However, note that the execution of promptForDishChoice and fetchPrices don't depend on the result of each other. While the user is choosing their dish, it's fine for the prices to be fetched in the background, but in the code above, the await operator causes the async function to pause until the choice is made, and then again until the prices are fetched. We can use Promise.all to run them concurrently, so that the user doesn't have to wait for the prices to be fetched before the result is given:

js
async function getPrice() {
  const [choice, prices] = await Promise.all([
    promptForDishChoice(),
    fetchPrices(),
  ]);
  return prices[choice];
}

Promise.all is the best choice of concurrency method here, because error handling is intuitive — if any of the promises reject, the result is no longer available, so the whole await expression throws.

Promise.all accepts an iterable of promises, so if you are using it to run several async functions concurrently, you need to call the async functions and use the returned promises. Directly passing the functions to Promise.all does not work, since they are not promises.

js
async function getPrice() {
  const [choice, prices] = await Promise.all([
    promptForDishChoice,
    fetchPrices,
  ]);
  // `choice` and `prices` are still the original async functions;
  // Promise.all() does nothing to non-promises
}

Promise.all fail-fast behavior

Promise.all is rejected if any of the elements are rejected. For example, if you pass in four promises that resolve after a timeout and one promise that rejects immediately, then Promise.all will reject immediately.

js
const p1 = new Promise((resolve, reject) => {
  setTimeout(() => resolve("one"), 1000);
});
const p2 = new Promise((resolve, reject) => {
  setTimeout(() => resolve("two"), 2000);
});
const p3 = new Promise((resolve, reject) => {
  setTimeout(() => resolve("three"), 3000);
});
const p4 = new Promise((resolve, reject) => {
  setTimeout(() => resolve("four"), 4000);
});
const p5 = new Promise((resolve, reject) => {
  reject(new Error("reject"));
});

// Using .catch:
Promise.all([p1, p2, p3, p4, p5])
  .then((values) => {
    console.log(values);
  })
  .catch((error) => {
    console.error(error.message);
  });

// Logs:
// "reject"

It is possible to change this behavior by handling possible rejections:

js
const p1 = new Promise((resolve, reject) => {
  setTimeout(() => resolve("p1_delayed_resolution"), 1000);
});

const p2 = new Promise((resolve, reject) => {
  reject(new Error("p2_immediate_rejection"));
});

Promise.all([p1.catch((error) => error), p2.catch((error) => error)]).then(
  (values) => {
    console.log(values[0]); // "p1_delayed_resolution"
    console.error(values[1]); // "Error: p2_immediate_rejection"
  },
);

Specifications

Specification
ECMAScript Language Specification
# sec-promise.all

Browser compatibility

BCD tables only load in the browser

See also