Bioptimus is a Paris-based AI biotech company building foundation models designed to learn and simulate human biology across every scale — from molecules and cells to tissues and whole organisms. Its flagship products include H-Optimus, the #1-ranked pathology foundation model trained on over one million histopathology slides, and M-Optimus, the first multimodal, multi-scale foundation model that jointly integrates pathology, spatial transcriptomics, and genomics into a single architecture. These models help pharma and clinical researchers extract deep biological insights from routine data without requiring new assays.

Bioptimus's models are trusted by 16 of the top 20 pharmaceutical companies and over 1,000 institutions worldwide, and have been downloaded more than 1.1 million times. The company operates an open-core commercialization model — H-Optimus-0 is freely available under an Apache 2.0 license, while newer models are sold via AWS Marketplace. The team was co-founded by alumni of Google DeepMind, Google Brain, and Owkin, and 60% of staff hold PhDs.

To scale its multimodal models, Bioptimus launched STELA (Spatial Tissue Embedding Learning Atlas), a multi-institutional initiative to build the world's largest clinically linked multimodal tissue atlas, targeting 100,000 deeply profiled patient specimens. Partnership agreements with institutions such as Cancer Research Horizons, FFCD, MIT, and 10x Genomics are expanding its proprietary data engine. Applications include drug target and biomarker discovery, indication expansion, clinical trial design, and treatment response prediction.