Open-source video models

Member lesson

Renting a GPU by the hour

Lesson 7 of 10 in Open-source video models.

Checked against the makers' documentation on 25 September 2026.

How to run code on a GPU in less than 10 lines of code · Modal
Getting started with Modal · Modal
How to Deploy a PyTorch GPU Pod on Runpod · Runpod
Lesson

Here are the rates. On 25 September 2026 Modal's per-second prices worked out to $0.80 an hour for an L4 (24 GB), $1.95 for an L40S (48 GB), $2.50 for an 80 GB A100, $3.03 for an RTX PRO 6000 (96 GB) and $3.95 for an H100, with $30 of free compute a month on its free Starter plan. RunPod's Pods, in its pricing table updated on 13 September, cost $0.34 an hour for an RTX 4090 on its Community Cloud or $0.74 on its Secure Cloud. An L40S there is $0.79 or $1.09, an H100 SXM $2.69 or $3.49. Hugging Face Inference Endpoints bill by the minute, at $1.80 an hour for an L40S and $2.50 for an A100 on AWS.

modalstart $run $idle 60 s $offrunpod podrunning $stopped: disk $deletedspace, gpustart $run $sleeps only if setyou stopped workingthe meter stops when the machine does, not you
When the meter runs: a Modal function bills while starting, running and sixty seconds idle; a RunPod pod bills while running and for its disk when stopped; a Space on a paid GPU bills until it sleeps, and only sleeps if you set it to.
An ink drawing of a taxi waiting at a kerb, its roof sign lit yellow, while a passenger with a bag walks away along the street.
A rented GPU is a taxi: the meter runs until the car is gone, not until you are.
Member

The rest of this is for members.

You have read the opening of this lesson. The rest of it comes with a membership.

One membership covers all four courses and Brands.

  • All lessons in the four Learn courses
  • All brand breakdowns in the Brands section
  • All quizzes, flashcards and crosswords
  • An AI teacher on every lesson, for up to 60 questions a day
  • All course material published during the membership

Price

Membership is from $12 a month on the Yearly plan, which is charged at $144 once a year. Lifetime access is $59, paid once.Membership is from ₹200 a month on the Yearly plan, which is charged at ₹2,400 + GST once a year. Lifetime access is ₹3,999 + GST, paid once.

Check what stuck

Members only

You can't take this check yet.

Members answer these 3 questions and keep the ticket for this lesson. Every lesson has one.

  1. Question 1, locked

    Which of these words means this? “A GPU rented per second only while your function runs, with the platform starting and stopping the machine for you.”

    • Serverless GPU
    • Release notes
    • Distilled checkpoint
    • Model CPU offload
  2. Question 2, locked

    Which of these words means this? “A whole rented machine with a GPU, billed per second from the moment it starts until you stop or delete it.”

    • export_to_video
    • ComfyUI-Manager
    • Excluded territory
    • Pod
  3. Question 3, locked

    Which of these words means this? “How long a serverless container waits idle before shutting down; Modal's default is 60 seconds, and you pay while it waits.”

    • Compute unit
    • Commit hash
    • Model card
    • Scale-down window

Elsewhere in the library

Hear about new courses.

Leave an email and we will tell you when something new lands. We do not write often.