Tempus Launches Paige Predict, Following $81M Paige AI Acquisition
Tempus has introduced Paige Predict, a new suite of AI-based digital pathology tools that analyze standard hematoxylin and eosin (H&E) stained tissue slides to estimate the likelihood of biomarker presence. The system is designed to guide molecular test ordering in cases where tumor tissue is limited and not sufficient for traditional next-generation sequencing (NGS) or immunohistochemistry (IHC) assays. Paige Predict uses AI to assess H&E slides and inform which molecular or IHC tests are most likely to return actionable results, helping optimize tissue use and reduce turnaround times.
Paige Predict addresses a common problem in clinical oncology—patients often have small or degraded biopsy samples that cannot support multiple rounds of molecular testing. This results in delays or missed opportunities to match patients with targeted therapies.
The foundation model from Paige, originally developed for digital pathology applications, was extended using this joint dataset following Tempus’s acquisition of Paige in 2025. Paige had previously developed the first FDA-cleared AI pathology tool, Paige Prostate, authorized for in vitro diagnosis of prostate cancer in needle biopsies or in vitro diagnosis, and in 2024 open-sourced its large-scale foundation models Virchow and PRISM, trained on over 2 million pathology slides and clinical reports, to support AI research in cancer diagnostics.
See also: 17 Companies Pioneering AI Foundation Models in Pharma and Biotech
Paige Predict covers predictions across 123 molecular biomarkers and oncogenic pathways in 16 cancer types, including non-small cell lung cancer (NSCLC), breast, prostate, colorectal, and pancreatic cancers. Predictions are delivered alongside routine pathology reports to support clinical decision-making in real time.
The underlying AI models were trained on a multimodal dataset combining digitized pathology images with matched molecular profiles from over 200,000 de-identified patient cases pooled from Tempus and Paige. The dataset includes structured genomic, clinical, and histological information and was used to build and validate the models for accuracy, robustness, and applicability across diverse populations.
See also: Tempus, AstraZeneca, and Pathos Partner to Build Oncology Foundation Model Using Multimodal Data
This launch expands Tempus’s goal to integrate AI into precision diagnostics and therapeutic decision-making. Recent efforts include the rollout of olivia, a personal health assistant for managing medical data, and the acquisition of Deep 6 AI to enhance clinical trial matching using natural language processing.
Tempus also partnered with Genialis to validate RNA-based biomarkers for targeted cancer therapy and launched Loop, a discovery platform combining real-world data, patient-derived organoids, and CRISPR screening. In the beginning of 2025, Tempus introduced xM for TRM, a liquid biopsy assay for monitoring response to immunotherapy in advanced cancers, and a year prior acquired Ambry Genetics for $600 million to expand into hereditary and clinical genetic testing.
Topic: AI in Bio
