Introduction

ImportError: No module named flaskext.sqlalchemy usually means the application is using an old Flask extension import style that no longer matches the installed package. Modern Flask-SQLAlchemy is imported from flask_sqlalchemy, and on Elastic Beanstalk you also need to make sure the dependency is actually installed in the deployed environment.

Why the old import breaks

Older Flask code sometimes used the flask.ext namespace.

# old style
from flaskext.sqlalchemy import SQLAlchemy

That style is legacy. Modern code should use:

from flask_sqlalchemy import SQLAlchemy

If your Elastic Beanstalk environment installs a current Flask-SQLAlchemy package, the legacy namespace import will fail even if the package itself is present.

Fix the application code first

Update all imports consistently.

1from flask import Flask
2from flask_sqlalchemy import SQLAlchemy
3
4application = Flask(__name__)
5application.config["SQLALCHEMY_DATABASE_URI"] = "sqlite:///app.db"
6db = SQLAlchemy(application)
7
8@application.route("/")
9def index():
10    return "ok"

Do not fix one file and forget helper modules or blueprints that still use the old namespace.

Make sure Elastic Beanstalk installs the packages

The second common cause is missing dependencies in requirements.txt.

1Flask==3.0.3
2Flask-SQLAlchemy==3.1.1
3SQLAlchemy==2.0.35
4gunicorn==22.0.0

Elastic Beanstalk installs packages from this file during deployment. If Flask-SQLAlchemy is missing or pinned to an incompatible version, the app may work locally and fail remotely.

Check the runtime environment

Elastic Beanstalk may use a different Python version or platform stack than your local development machine.

Useful commands:

eb platform show
eb logs

The log output often makes it clear whether the issue is:

• missing package installation
• bad import path
• incompatible Python version
• startup failure after deployment

Keep the entrypoint simple

A minimal EB-friendly app structure reduces confusion during debugging.

1# application.py
2from flask import Flask
3from flask_sqlalchemy import SQLAlchemy
4
5application = Flask(__name__)
6application.config["SQLALCHEMY_DATABASE_URI"] = "sqlite:///app.db"
7application.config["SQLALCHEMY_TRACK_MODIFICATIONS"] = False
8
9db = SQLAlchemy(application)
10
11@application.route("/")
12def home():
13    return "Elastic Beanstalk is working"

Then start it with Gunicorn using the expected object path.

gunicorn application:application

That gives EB a predictable app target.

Why this often appears only in deployment

Developers sometimes keep older packages installed locally or have a working virtual environment that hides the outdated import. Deployment exposes the problem because the environment is built fresh from declared dependencies.

That difference between local state and declared state is exactly why Elastic Beanstalk can reveal packaging mistakes that never showed up during casual local testing.

That is actually useful. It forces the project to rely on explicit requirements rather than accidental local state.

A practical debugging checklist

When this error appears on Elastic Beanstalk:

Treat it as both an application-code problem and a deployment-definition problem. The import path may be wrong, but the packaged environment may also be missing the extension entirely.

1. replace all flaskext.sqlalchemy imports with flask_sqlalchemy
2. confirm Flask-SQLAlchemy is present in requirements.txt
3. redeploy and inspect EB logs
4. verify Python version and package compatibility
5. confirm the application entrypoint is correct

In most cases, steps one and two solve it.

Common Pitfalls

A common mistake is fixing the import in one file but leaving legacy imports elsewhere in the project.

Another mistake is assuming the package exists remotely just because it exists locally.

A third mistake is debugging Flask routing or Gunicorn startup before fixing the import path and dependency declaration, which are usually the real issue.

Summary

• 'flaskext.sqlalchemy is a legacy import style.'
• Modern Flask-SQLAlchemy should be imported from flask_sqlalchemy.
• Elastic Beanstalk must also install the package through requirements.txt.
• Check EB logs and platform runtime if local and deployed behavior differ.
• Fix imports first, then verify dependency installation and app entrypoint.