This project aims to classify the bias in news media in real time. With the help of labeled data created by open source projects opensources.co and mediabiasfactcheck.com which document bias in thousands of news sources, a model of what makes a source biased is constructed to classify new information. This project periodically collects tens of thousands of articles from the labeled news sources and trains a custom-built neural network on the articles in order to model and characterize bias. When a user visits this site and submits a news website url for analysis, a system of EC2 instances and AWS Lambda functions gathers a few dozen of the latest articles from the site. The collected text is sent to the neural network model residing in an AWS Lambda function, and the results are rendered in matplotlib and published via flask. This project is under continued development as UX, data visualization and modeling are expanded and refined.