Skip to main content

Access to SAP RPT Model

Steps

  1. Overview
  2. Pre-requisites
  3. Key Choices and Guidelines
  4. Implementation

1 Overview​

Description​

SAP-RPT-1 (Relational Pretrained Transformer) is a table-native foundation model designed specifically for tabular and relational data. It performs classification and regression tasks without any training or fine-tuning through in-context learning. Unlike traditional ML approaches, RPT-1 learns patterns from example rows provided directly in the API request and predicts values for query rows.

Expected Outcome​

Provide instant, high-quality predictions on structured business data without the need for training infrastructure, datasets, or model fine-tuning cycles. Applications can leverage RPT-1 to automate classification and regression tasks on enterprise tabular data.