FairBench is an AI fairness/bias assessment library that aims to present a multifaceted view of various literature considerations. It does so by splitting standard definitions of fairness in basic buiilding blocks and re-combines the blocks to create a wide range of definitions that catter to a wide range of real-world context. This enables human participation in auditing AI by providing many prospective biases. FairBench also allows its users to backtrack computations to find the root causes of discrimination, as well as compare multiple algorithms and track their evolution over time. It also provides several console-based, static, and interactive visualization engines to look at fairness reports with different levels of conciseness vs detail. Finally, it includes investigative post-processing of fairness reports, including the option to add caveats and recommendation when trying to reuse populat practices to generate fairness model cards.