What does one Remotion Lambda render cost?
Remotion published four measured renders, all at the default 2048 MB of memory, 10 GB disk, default concurrency, in us-east-1, on Remotion 4.0.381 (Remotion cost example):
| Render | Warm Lambda | Cold Lambda |
|---|---|---|
| Hello World template | $0.001 (7.56 s) | $0.001 (11.02 s) |
| 1-minute video from the same S3 bucket | $0.017 (18.91 s) | $0.021 (15.52 s) |
| 10-minute remote HD video | $0.103 (56.09 s) | $0.108 (60.98 s) |
| 10-second remote 4K video | $0.013 (45.28 s) | $0.014 (53.09 s) |
Lambda compute only. Remotion lists the extras: S3 egress for pulling assets, S3 storage, CloudWatch logs and, for teams of 4+, the Remotion license. Remotion recommends measuring your own composition; region, bundle size and latency all move the number.
What does a month of renders cost?
Here are Remotion's cold-start figures (the higher of the two) multiplied by monthly volume. The "company" columns add Remotion for Automators: $0.01 per render, never less than $100 a month (remotion.dev/license).
1-minute video from the same S3 bucket
| Renders a month | AWS compute | Total, team of up to 3 | Total, company of 4+ | Per render, company |
|---|---|---|---|---|
| 100 | $2.10 | $2.10 | $102 | $1.02 |
| 1,000 | $21.00 | $21.00 | $121 | $0.12 |
| 10,000 | $210 | $210 | $310 | $0.031 |
| 100,000 | $2,100 | $2,100 | $3,100 | $0.031 |
10-minute remote HD video
| Renders a month | AWS compute | Total, team of up to 3 | Total, company of 4+ | Per render, company |
|---|---|---|---|---|
| 100 | $10.80 | $10.80 | $111 | $1.11 |
| 1,000 | $108 | $108 | $208 | $0.21 |
| 10,000 | $1,080 | $1,080 | $1,180 | $0.12 |
| 100,000 | $10,800 | $10,800 | $11,800 | $0.12 |
10-second remote 4K video
| Renders a month | AWS compute | Total, team of up to 3 | Total, company of 4+ | Per render, company |
|---|---|---|---|---|
| 100 | $1.40 | $1.40 | $101 | $1.01 |
| 1,000 | $14.00 | $14.00 | $114 | $0.11 |
| 10,000 | $140 | $140 | $240 | $0.024 |
| 100,000 | $1,400 | $1,400 | $2,400 | $0.024 |
Our arithmetic on Remotion's published per-render costs; linear scaling, no AWS free tier subtracted, no S3, CloudWatch or bandwidth. At low volume the $100 license minimum is most of the bill; at high volume, AWS is.
How is AWS Lambda priced?
AWS charges per request and per GB-second (memory × run time). From the AWS Lambda pricing page, US East (N. Virginia): $0.2 per million requests; $0.0000166667 per GB-second on x86 and $0.0000133334 on Arm, in the first tier. The free tier covers 1,000,000 requests and 400,000 GB-seconds a month. Remotion's Lambda runtime is built for ARM64 (runtime docs), and because cost scales with memory, Remotion says reducing memory "will linearly decrease the cost" (optimizing for cost).
How does Remotion Lambda work?
- A Lambda function and an S3 bucket are created in your AWS account; your Remotion project is deployed to S3 as a website.
- One Lambda opens the project and starts many others in parallel, each rendering a small part of the video.
- The first Lambda stitches the chunks together and uploads the finished video to S3.
Setup per Remotion's guide: npx remotion add @remotion/lambda, an IAM policy named exactly remotion-lambda-policy (from npx remotion lambda policies role), a role named remotion-lambda-role, and a user with an access key in your .env. Then npx remotion lambda functions deploy, npx remotion lambda sites create and npx remotion lambda render [comp-id].
What are the limits?
- Videos up to about 80 minutes at Full HD, before AWS's 15-minute timeout.
- Up to 10 GB of storage per function, so output of about 5 GB (around 2 hours of Full HD).
- 1,000 concurrent Lambdas per region by default (AWS can raise it).
- No AV1 encoding on Lambda. 25 AWS regions supported.
@remotion/lambda docs, November 10, 2026.
When is Lambda the wrong choice?
Remotion recommends Lambda for most people for speed, setup and scale, but its own comparison says Lambda compute "is more expensive than other options": Cloud Run is cheaper because there is no distributed overhead, and a server has by far the cheapest compute if you keep it busy. Vercel Sandbox is the simplest if you already use Vercel. HeyGen's HyperFrames also renders on AWS Lambda, Cloud Run or HeyGen's own cloud, under Apache 2.0 with no license fee, though HeyGen calls its Lambda path newer than Remotion's (Remotion vs HyperFrames).
Sources
- How much does Remotion Lambda cost?, @remotion/lambda, Setup, Optimizing for cost, SSR comparison: checked November 10, 2026
- Remotion License & Pricing: checked November 10, 2026
- AWS Lambda pricing: checked November 10, 2026
