Skip to main content

Setup

import Tabs from '@theme/Tabs'; import TabItem from '@theme/TabItem'; import { YouTube } from "../../components/YouTube";

1. Install @remotion/lambda​

<Tabs defaultValue="npm" values={[ { label: 'npm', value: 'npm', }, { label: 'yarn', value: 'yarn', }, { label: 'pnpm', value: 'pnpm', }, ] }> 

bash
npm i @remotion/lambda
bash
npm i @remotion/lambda
bash
pnpm i @remotion/lambda
bash
pnpm i @remotion/lambda
 
bash
yarn add @remotion/lambda
bash
yarn add @remotion/lambda

Also update all the other Remotion packages to have the same version: remotion, @remotion/cli and others.

note

Make sure no package version number has a ^ character in front of it as it can lead to a version conflict.

Your package.json should look like the following:

json
  "@remotion/cli": "3.0.0", // Replace 3.0.0 with the current version; install as a dev dependency
  "@remotion/lambda": "3.0.0", // Remove any `^` character
  // ...
  "remotion": "3.0.0",
json
  "@remotion/cli": "3.0.0", // Replace 3.0.0 with the current version; install as a dev dependency
  "@remotion/lambda": "3.0.0", // Remove any `^` character
  // ...
  "remotion": "3.0.0",

If any of the above is missing, add them to the package.json file and then run npm install, yarn install or pnpm install.

2. Create role policy​

  • Go to AWS account IAM Policies section
  • Click on "Create policy"
  • Click on JSON
  • In your project, type npx remotion lambda policies role in the command line and copy it into the "JSON" field on AWS.
  • Click next. On the tags page, you don't need to fill in anything. Click next again.
  • Give the policy exactly the name remotion-lambda-policy. The other fields can be left as they are.

3. Create a role​

  • Go to AWS account IAM Roles section
  • Click "Create role".
  • Under "Use cases", select "Lambda". Click next.
  • Under "Permissions policies", filter for remotion-lambda-policy and click the checkbox to assign this policy. Click next.
  • In the final step, name the role remotion-lambda-role exactly. You can leave the other fields as is.
  • Click "Create role" to confirm.

4. Create a user​

  • Go to AWS account IAM Users section
  • Click Add users
  • Enter any username, such as remotion-user.
  • Don't check the "Enable console access" option. You don't need it.
  • Click "Next".
  • Click "Next" again without changing any settings. You should now be on the "Review and Create" step.
  • Click "Create user".

5. Create an access key for the user​

  • Go to AWS account IAM Users section
  • Click on the name of the user that was created in step 4.
  • Navigate to the "Security Credentials" tab, and scroll down to the "Access Keys" section.
  • Click the "Create access key" button.
  • Select "Application running on an AWS compute service".
  • Ignore warnings that might appear and check the "I understand the recommendations..." checkbox.
  • Click "Next".
  • Click "Create access key".
  • Add a .env file to your project's root and add the credentials you just copied in the following format:
.env
txt
REMOTION_AWS_ACCESS_KEY_ID=<Access key ID>
REMOTION_AWS_SECRET_ACCESS_KEY=<Secret access key>
.env
txt
REMOTION_AWS_ACCESS_KEY_ID=<Access key ID>
REMOTION_AWS_SECRET_ACCESS_KEY=<Secret access key>

6. Add permissions to your user​

  • Go to AWS account IAM Users section
  • Select the user you just created.
  • Click "Add inline policy" under the "Add Permissions" dropdown in the "Permissions policies" panel.
  • Click the tab "JSON".
  • Enter in your terminal: npx remotion lambda policies user and copy into the AWS text field what gets printed.
  • Click "Review policy".
  • Give the policy a name. For example remotion-user-policy, but it can be anything.
  • Click "Create policy" to confirm.

7. Optional: Validate the permission setup​

Check all user permissions and validate them against the AWS Policy simulator by executing the following command:

bash
npx remotion lambda policies validate
bash
npx remotion lambda policies validate

For the following steps, you may execute them on the CLI, or programmatically using the Node.JS APIs.

8. Deploy a function​

<Tabs defaultValue="cli" values={[ { label: 'CLI', value: 'cli', }, { label: 'Node.JS', value: 'node', }, ] }>

Deploy a function that can render videos into your AWS account by executing the following command:

bash
npx remotion lambda functions deploy
bash
npx remotion lambda functions deploy

You can deploy a function that can render videos into your AWS account using deployFunction().

ts
const { functionName } = await deployFunction({
  region: "us-east-1",
  timeoutInSeconds: 120,
  memorySizeInMb: 2048,
  createCloudWatchLogGroup: true,
});
ts
const { functionName } = await deployFunction({
  region: "us-east-1",
  timeoutInSeconds: 120,
  memorySizeInMb: 2048,
  createCloudWatchLogGroup: true,
});

The function name is returned which you'll need for rendering.

The function consists of necessary binaries and JavaScript code that can take a serve URL and make renders from it. A function is bound to the Remotion version, if you upgrade Remotion, you need to deploy a new function. A function does not include your Remotion code, it will be deployed in the next step instead.

9. Deploy a site​

<Tabs defaultValue="cli" values={[ { label: 'CLI', value: 'cli', }, { label: 'Node.JS', value: 'node', }, ] }>

Run the following command to deploy your Remotion project to an S3 bucket. Pass as the last argument the entry point of the project.

bash
npx remotion lambda sites create src/index.ts --site-name=my-video
bash
npx remotion lambda sites create src/index.ts --site-name=my-video

A serveUrl will be printed pointing to the deployed project.

When you update your Remotion video in the future, redeploy your site. Pass the same --site-name to overwrite the previous deploy. If you don't pass --site-name, a unique URL will be generated on every deploy.

First, you need to create an S3 bucket in your preferred region. If one already exists, it will be used instead:

ts
import path from "path";
import { deploySite, getOrCreateBucket } from "@remotion/lambda";
 
const { bucketName } = await getOrCreateBucket({
  region: "us-east-1",
});
ts
import path from "path";
import { deploySite, getOrCreateBucket } from "@remotion/lambda";
 
const { bucketName } = await getOrCreateBucket({
  region: "us-east-1",
});

Next, upload your Remotion project to an S3 bucket. Specify the entry point of your Remotion project, this is the file where registerRoot() is called.

ts
const { serveUrl } = await deploySite({
  bucketName,
  entryPoint: path.resolve(process.cwd(), "src/index.ts"),
  region: "us-east-1",
  siteName: "my-video",
});
ts
const { serveUrl } = await deploySite({
  bucketName,
  entryPoint: path.resolve(process.cwd(), "src/index.ts"),
  region: "us-east-1",
  siteName: "my-video",
});

When you update your Remotion video in the future, redeploy your site. Pass the same siteName to overwrite the previous deploy. If you don't pass siteName, a unique URL will be generated on every deploy.

10. Check AWS concurrency limit​

Check the concurrency limit that AWS has given to your account:

npx remotion lambda quotas
npx remotion lambda quotas

By default, it is 1000 concurrent invocations per region. However, new accounts might have a limit as low as 10. Each Remotion render may use as much as 200 functions per render concurrently, so if your assigned limit is very low, you might want to request an increase right away.

11. Render a video​

<Tabs defaultValue="cli" values={[ { label: 'CLI', value: 'cli', }, { label: 'Node.JS', value: 'node', }, ] }>

Take the URL you received from the step 9 - your "serve URL" - and run the following command. Also pass in the ID of the composition you'd like to render.

bash
npx remotion lambda render <serve-url> <composition-id>
bash
npx remotion lambda render <serve-url> <composition-id>

Progress will be printed until the video finished rendering. Congrats! You rendered your first video using Remotion Lambda 🚀

You already have the function name from a previous step. But since you only need to deploy a function once, it's useful to retrieve the name of your deployed function programmatically before rendering a video in case your Node.JS program restarts. We can call getFunctions() with the compatibleOnly flag to get only functions with a matching version.

ts
import {
  getFunctions,
  renderMediaOnLambda,
  getRenderProgress,
} from "@remotion/lambda";
 
const functions = await getFunctions({
  region: "us-east-1",
  compatibleOnly: true,
});
 
const functionName = functions[0].functionName;
ts
import {
  getFunctions,
  renderMediaOnLambda,
  getRenderProgress,
} from "@remotion/lambda";
 
const functions = await getFunctions({
  region: "us-east-1",
  compatibleOnly: true,
});
 
const functionName = functions[0].functionName;

We can now trigger a render using the renderMediaOnLambda() function.

ts
const { renderId, bucketName } = await renderMediaOnLambda({
  region: "us-east-1",
  functionName,
  serveUrl: url,
  composition: "HelloWorld",
  inputProps: {},
  codec: "h264",
  imageFormat: "jpeg",
  maxRetries: 1,
  framesPerLambda: 20,
  privacy: "public",
});
ts
const { renderId, bucketName } = await renderMediaOnLambda({
  region: "us-east-1",
  functionName,
  serveUrl: url,
  composition: "HelloWorld",
  inputProps: {},
  codec: "h264",
  imageFormat: "jpeg",
  maxRetries: 1,
  framesPerLambda: 20,
  privacy: "public",
});

The render will now run and after a while the video will be available in your S3 bucket. You can at any time get the status of the video render by calling getRenderProgress().

ts
while (true) {
  await new Promise((resolve) => setTimeout(resolve, 1000));
  const progress = await getRenderProgress({
    renderId,
    bucketName,
    functionName,
    region: "us-east-1",
  });
  if (progress.done) {
    console.log("Render finished!", progress.outputFile);
    process.exit(0);
  }
  if (progress.fatalErrorEncountered) {
    console.error("Error enountered", progress.errors);
    process.exit(1);
  }
}
ts
while (true) {
  await new Promise((resolve) => setTimeout(resolve, 1000));
  const progress = await getRenderProgress({
    renderId,
    bucketName,
    functionName,
    region: "us-east-1",
  });
  if (progress.done) {
    console.log("Render finished!", progress.outputFile);
    process.exit(0);
  }
  if (progress.fatalErrorEncountered) {
    console.error("Error enountered", progress.errors);
    process.exit(1);
  }
}

This code will poll every second to check the progress of the video and exit the script if the render is done. Congrats! Check your S3 Bucket - you just rendered your first video using Remotion Lambda 🚀

Next steps​