ModuleNotFoundError: No Module Named ‘anthropic’ (2026)

Anthropic is the official Python SDK for Claude. If your script raises ModuleNotFoundError: No module named ‘anthropic’, install it with one pip command. Make sure your environment also has the ANTHROPIC_API_KEY environment variable set.

Step 1: Install the SDK

pip install anthropic

# Or with bedrock support (AWS hosted Claude):
pip install 'anthropic[bedrock]'

# With vertex (Google Cloud hosted Claude):
pip install 'anthropic[vertex]'

Step 2: Set your API key

# Linux / macOS
export ANTHROPIC_API_KEY="sk-ant-..."

# Windows PowerShell
$env:ANTHROPIC_API_KEY = "sk-ant-..."

# Or in Python (less secure, do not commit):
import os
os.environ['ANTHROPIC_API_KEY'] = 'sk-ant-...'

Step 3: First request

from anthropic import Anthropic

client = Anthropic()  # reads ANTHROPIC_API_KEY from env

message = client.messages.create(
    model="claude-opus-4-7",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello, Claude!"}]
)
print(message.content[0].text)

Step 4: Streaming response

with client.messages.stream(
    model="claude-opus-4-7",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Write a haiku"}]
) as stream:
    for text in stream.text_stream:
        print(text, end="", flush=True)

Why this error happens

Debugging checklist for ModuleNotFoundError: anthropic

• Confirm which Python is running: which python and python --version.
• Confirm which pip installed the package: python -m pip show anthropic.
• If you are in a virtual env, confirm it is activated. (venv) should appear in your shell prompt.
• If Jupyter, confirm the kernel matches. Restart the kernel after install.

Correct install commands for common setups

# Standard Python (system or venv)
python -m pip install anthropic

# Poetry
poetry add anthropic

# Conda environment
conda run -n myenv pip install anthropic

# Jupyter Notebook (install into the running kernel)
import sys
!{sys.executable} -m pip install anthropic

# uv (fastest 2026 workflow)
uv add anthropic

Common mistakes with the Anthropic SDK

• Installing to global Python while running in a venv. Use python -m pip install instead of raw pip install.
• Wrong package name. The Anthropic Python SDK is anthropic, not anthropic-ai or claude-sdk.
• Corporate proxy blocking PyPI. Set HTTPS_PROXY or use --index-url.
• Old cached wheel. Force a clean install with pip install --no-cache-dir --upgrade anthropic.

Verify the install and run a smoke test

import anthropic
print(anthropic.__version__)

# Smoke test (needs ANTHROPIC_API_KEY set)
client = anthropic.Anthropic()
msg = client.messages.create(
    model="claude-opus-4-8",
    max_tokens=64,
    messages=[{"role": "user", "content": "Say hi"}]
)
print(msg.content[0].text)

If the import prints a version but the API call fails, the install is fine. Your problem is the API key or network, not the module.

Real-world example: Anthropic SDK in a FastAPI service

Here is a full working example that runs the Anthropic SDK inside a FastAPI service. If your import works but you still hit runtime errors, this is the reference to compare against.

# main.py
from fastapi import FastAPI
from pydantic import BaseModel
import anthropic

app = FastAPI()
client = anthropic.Anthropic()

class ChatRequest(BaseModel):
    message: str

@app.post("/chat")
def chat(req: ChatRequest):
    reply = client.messages.create(
        model="claude-opus-4-8",
        max_tokens=512,
        messages=[{"role": "user", "content": req.message}]
    )
    return {"reply": reply.content[0].text}

# Install:  uv add anthropic fastapi uvicorn
# Run:      uvicorn main:app --reload
# Test:     curl -X POST http://127.0.0.1:8000/chat -H "content-type: application/json" -d '{"message":"Say hi"}'

If the FastAPI process fails with ModuleNotFoundError even after uv add anthropic, the most common cause is running uvicorn from a different Python. Use uv run uvicorn main:app --reload to force the correct interpreter.

Related ModuleNotFoundError patterns

• No module named “openai”: same fix pattern. Install with uv add openai or python -m pip install openai.
• No module named “langchain”: the meta package. You often want a specific submodule like langchain-anthropic or langchain-openai.
• No module named “dotenv”: the package name on PyPI is python-dotenv, not dotenv. This one bites everyone once.

Quick reference: environment-specific install commands

Use this as a fast lookup when the ModuleNotFoundError comes back weeks or months later. The install command changes based on how the project is bootstrapped, and using the wrong one is why the same error keeps showing up on the same laptop across projects.

• Plain venv or system Python: python -m pip install anthropic. Always use python -m pip, never bare pip, so you install into the interpreter that is actually running your script.
• uv-managed project (recommended in 2026): uv add anthropic and run scripts with uv run python your_script.py. The runner picks the right virtual env automatically.
• Poetry project: poetry add anthropic, then poetry run python your_script.py.
• Conda environment: activate the env first with conda activate myenv, then pip install anthropic. Do not mix conda install and pip install for the same package.
• Jupyter Notebook: install into the running kernel with %pip install anthropic (the magic auto-picks the correct interpreter). Restart the kernel afterward.
• Docker image: add RUN pip install anthropic to your Dockerfile, then rebuild. Do not install at runtime.

The single most common cause of the module error coming back is running your entry point with a different Python than the one you installed into. Verify with python -c "import anthropic; print(anthropic.__file__)" to see the actual install path.

Adrian Mercurio

Full-Stack Developer at PIES IT Solution

Specializes in building complete capstone projects with full documentation. Strong background in PHP/MySQL development and database design. Has personally built and tested over 30 capstone-ready projects with ER diagrams, DFDs, and chapter-by-chapter thesis documentation.

Expertise: PHP · Laravel · Database Design · Capstone Projects · C# · C · C++ · Python · AI Projects
 · View all posts by Adrian Mercurio →