Deepseek took the world by suprise. Not just by its performance but also by its accessabilty Making deepseek models open source is a great step to involve more people in its development and also making accessable for devs
In this post I will show how to run locally
While you can run it locally you still would need a powerful PC and you may not be able to run the models with the high parameters number
It may slow down your machine and be slow as well and make it unusable with other application or in general So most of the issue you will be facing are on the hardware side
First you will need to install Ollama
Installation instruction is in the Link
Then run in your command line
ollama run deepseek-r1:14b
The number 14b at the end indicates the number of parameters depending on how well your hardware would perform you will able to decide which version would be suitable
You can either use the command prompt chat or make Http requests to your local ollama server
curl http://localhost:11434/api/generate -d '{ "model": "deepseek-r1:14b", "prompt": "What year were you born ?" }'Enjoy!