Python Error: ModuleNotFoundError — No Module Named

Traceback (most recent call last):
  File "/Users/dev/project/app.py", line 3, in <module>
    import requests
ModuleNotFoundError: No module named 'requests'

ModuleNotFoundError is Python’s clearest error. The interpreter looked for a module on sys.path, didn’t find it, and stopped. The fix is almost always one of two things: you installed into a different Python than the one running your code, or your local code isn’t on sys.path. Both are diagnosable in 30 seconds with python -m pip list and python -c 'import sys; print(sys.path)'.

The single rule that prevents 90% of Python environment pain: always use python -m pip instead of bare pip. It guarantees the install and the run use the same interpreter, eliminating the most common variant of this error.

Why this happens

• Wrong virtualenv (or no venv at all). You activated a venv, then opened a new shell that didn't inherit it. Or your IDE runs the system Python while your terminal uses the project's venv. `pip install requests` went to one Python; `python app.py` ran a different one.
• pip vs python version mismatch. You ran `pip install` (which points to one Python, often the system one) but invoked `python3` (a different one). On macOS especially, `python` may be 3.9 (Apple-shipped) and `python3` may be 3.11 (Homebrew). The fix is `python -m pip install <pkg>` — that always uses the same Python.
• Project not installed editably / wrong PYTHONPATH. Your imports use `from myapp.utils import …` but `myapp` is your local code, not on sys.path. Either the project isn't `pip install -e .`'d, or you're running from a directory that doesn't have `myapp` as a subfolder.
• Missing __init__.py in a package directory. In Python 3.3+ namespace packages don't need `__init__.py`, but mixing namespace and regular packages, or having a stray empty directory, breaks imports. If `myapp/utils.py` works but `from myapp.utils import x` doesn't, look for missing `__init__.py`.
• Module installed but for a different Python or platform. You're inside Docker / a container with a different Python build (alpine vs slim, 3.10 vs 3.12) than your host. The host has `requests` installed; the container doesn't. Or you copied `site-packages` between architectures and binary wheels (numpy, pandas) don't load.

How to fix it

Fixes are ordered by likelihood. Start with the first one that matches your context.

# Confirm which Python you'll actually run:
which python
python --version

# Install into THAT interpreter, not the system one:
python -m pip install requests

# Verify it's actually importable from that Python:
python -c "import requests; print(requests.__version__)"

# Create + activate a venv (Python 3.11+):
python -m venv .venv
source .venv/bin/activate         # macOS/Linux
# .venv\Scripts\activate          # Windows PowerShell

# Install your dependencies:
python -m pip install -r requirements.txt

# Confirm the venv is active:
which python      # should print .../.venv/bin/python
python -m pip list

[project]
name = "myapp"
version = "0.1.0"
requires-python = ">=3.10"
dependencies = [
  "requests>=2.31",
  "pydantic>=2",
]

[build-system]
requires = ["setuptools>=68"]
build-backend = "setuptools.build_meta"

[tool.setuptools.packages.find]
where = ["src"]

import sys, os, site
print("python:", sys.executable)
print("version:", sys.version)
print("cwd:", os.getcwd())
print("PYTHONPATH:", os.environ.get('PYTHONPATH'))
print("site-packages:")
for p in site.getsitepackages():
    print("  ", p)
print("sys.path:")
for p in sys.path:
    print("  ", p)

Detection and monitoring in production

Add `python -c "import myapp; import otherdeps"` as a smoke test in CI on every PR — ModuleNotFoundError is deterministic and trivially catchable. For Docker images, run the smoke test as a `RUN` in the build (not just at container startup) so a missing dep fails the build, not the deploy. Track ModuleNotFoundError in production logs by exception class; in healthy services it should be zero.