CrewAI Installation Using Anaconda

In this article CrewAI Installation Using Anaconda.

Install anaconda from Anaconda Download.

Start powershell from anaconda folder ( similar to below)

We’ll create a new conda environment for crewai. Not a necessary steps but for a safety net, not impacting our existing ongoing conda envionment or python development work due to any dependencies packages or libraries installation.
Give command conda create -n crewai-handson to create new environment called crewai-handson
In case you are looking for specific version of python you can mention same - conda create -n crewai-handson python=3.11.9

Now activate this new environment using command - conda activate crewai-handson

I am using cursor ide for development of my crewai projects. Cursor-AI ( build on VS code) have it own terminal where you can activate your conda environment or use the anaconda shell in separate command or terminal window. Your choice of convenience.

If your cursor or windows command is not able to recognise conda command then add conda location to your environment variable in path.

Either in powershell/command prompt/cursor terminal, execute command conda activate crewai-handson
If not done sucessfully as Step#5

Cursor Terminal

Windows Powershell Command Prompt

Anaconda Powershell Command Prompt

Now let’s install packages required for crewai using pip - pip install crewai 'crewai[tools]' if pip is not installed, install pip using command - conda install pip

One of the cool feature of crewai is that it gives you a boilerplate code using crewai command so that you don’t have to create a skeleton code.
You can run the command - crewai create crew "ashish-sample-crewai" and it will create a crewai project name ashish-sample-crewai and with all the basic skeleton code or boilerplat code withhin it.

By default crewai uses OpenAI api. However in above crewai projection creation, I used the Deepseek LLM so I need to add environment files and give the my DeepSeek LLM api keys. Else it will try to connect to openAI LLM and you will get authentication error as you have not mentioned OpenAI API keys.

Congratulation, you have created crewai project successfully.

If you want to run this simply give command - crewai run Crewai will do the build and your crew will run

I received the OpenAI api error as expected because I didn’t provide api keys.
However it shows crewai set-up is correct and working as desired.