Documentation

Everything you need to get started with ArbitrAI — from uploading your first document to interpreting evaluation results.

Core User Flows

Follow these guides to learn the key workflows in ArbitrAI.

Learn how to upload documents, organize them into datasets, and prepare for evaluation.

Create ground truth labels using the visual annotation tool — draw bounding boxes, assign keys and values, and use the auto labeler.

Walk through the four-step wizard to configure and launch a benchmark evaluation run.

Understand evaluation results through the aggregate comparison table, metrics configuration, and detailed per-document model output views.

Looking for specifics? Browse the detailed reference for each page in the platform.