Getting Started

Calibr API Documentation

Deploy scorecards built in Calibr and integrate real-time credit risk scoring into any application. Our REST API lets you score applicants, manage model deployments, and run champion/challenger experiments with a few lines of code.

Base URL

All API requests are made to the following base URL. Every endpoint described in this documentation is relative to this base.

Quick Links

How It Works

Calibr is a desktop application for building, validating, and exporting credit risk scorecards. Once you are satisfied with a model, you deploy it to the Calibr API. From that point, your application sends applicant data to the scoring endpoint and receives a score, risk grade, and point breakdown in the response.

The API supports single-applicant scoring for real-time decisions and batch scoring for portfolio reviews. You can run multiple model versions simultaneously using champion/challenger deployments and control traffic splits via the API.

Key Concepts

Models — A scorecard exported from the Calibr desktop app. Each model has a versioned artifact containing bins, weights, and intercepts.

Deployments — A live endpoint backed by one or more model versions. Deployments have a unique slug used in the scoring URL.

Scoring — The act of sending applicant features to a deployment and receiving a numerical score, a risk grade, and a point-level explanation.

API Keys — Bearer tokens scoped to your organization. Keys can be created, listed, and revoked via the API or the Calibr dashboard.