How to beginner · 3 min read

Modal free tier explained

Q: Modal free tier explained

The Modal platform offers a free tier that includes limited GPU hours and CPU usage for running serverless functions with GPU acceleration. You can start using the free tier by signing up and setting your API key in the environment, then deploying Python functions with GPU support using modal.App and @app.function(gpu=...) decorators.

Quick answer

The Modal platform offers a free tier that includes limited GPU hours and CPU usage for running serverless functions with GPU acceleration. You can start using the free tier by signing up and setting your API key in the environment, then deploying Python functions with GPU support using modal.App and @app.function(gpu=...) decorators.

PREREQUISITES

Python 3.8+
Modal account with free tier access
pip install modal

Setup

Install the modal Python package and set your Modal API key as an environment variable to access the free tier.

bash

pip install modal

output

Collecting modal
  Downloading modal-1.x.x-py3-none-any.whl (xx kB)
Installing collected packages: modal
Successfully installed modal-1.x.x

Step by step

Use the free tier by writing a simple Python function with GPU support and deploying it with Modal.

python

import os
import modal

# Set your Modal API key in environment variable before running
# export MODAL_API_KEY=os.environ["MODAL_API_KEY"]

app = modal.App("my-free-tier-app")

@app.function(gpu="A10G", image=modal.Image.debian_slim().pip_install("torch"))
def run_inference(prompt: str) -> str:
    import torch
    # Dummy inference example
    return f"Processed prompt: {prompt}"

if __name__ == "__main__":
    with modal.runner.deploy_stub(app):
        result = run_inference.remote("Hello from Modal free tier")
        print(result)

output

Processed prompt: Hello from Modal free tier

Common variations

You can run CPU-only functions by omitting the gpu parameter or use different GPU types if available in your free tier. Async functions and web endpoints are also supported.

python

import modal

app = modal.App("my-app")

# CPU-only function
@app.function()
def cpu_task(x: int) -> int:
    return x * x

# Async function example
@app.function()
async def async_task(msg: str) -> str:
    return f"Async received: {msg}"

# Web endpoint example
@app.function().web_endpoint(method="POST")
def echo(request):
    data = request.json()
    return {"echo": data.get("message", "")}

Troubleshooting

If you see authentication errors, ensure MODAL_API_KEY is set correctly in your environment.
If GPU resources are unavailable, check your free tier usage limits or upgrade your plan.
For deployment issues, verify your function decorators and Modal app name are correct.

✅

Key Takeaways

Modal free tier provides limited GPU and CPU hours for serverless function deployment.
Set MODAL_API_KEY environment variable to authenticate your Modal client.
Use @app.function(gpu="A10G") decorator to run GPU workloads on the free tier.
You can deploy CPU-only, async, and web endpoint functions within the free tier limits.

Verified 2026-04

Verify ↗