bioptimus
manifesto

M-Optimus: Revolutionizing Drug Discovery and Development with a World Model of Biology 
Executive summary
The opportunity
The solution
Strategic impact
The roadmap
The data
The journey
References

Executive summary

Bioptimus is revolutionizing the historically costly and failure-prone drug discovery and clinical development process by launching M-Optimus, a universal World Model of Biology. This pioneering platform disrupts the current knowledge paradigms, which are replete with fragmented and siloed data spanning complex biological systems; harmonizing them within an unbiased, multimodal, multi-scale, universal foundation model capable of delivering patient-specific precision insights.

By leveraging the explosion of multimodal data and advanced AI, M-Optimus is a unifying foundation model that can underpin every stage of biological discovery and development, modeling robust representations of cells, tissues and patients across diseases and populations, providing a universal framework for understanding biology, predicting clinical outcomes, and guiding therapeutic decisions.

M-Optimus learns interactions across all scales and modalities (omics, images, clinical records). At its core, M-Optimus is comprised of universal embeddings of biomedical data and counterfactual simulations, enabling researchers to predict intervention outcomes in silico, significantly de-risking clinical trials (projected to boost asset rNPV by 20−30%), accelerating discovery, R&D timelines and powering the emergence of synthetic trial arms, multiscale digital twins, and lucrative preventive healthcare strategies. 

Bioptimus's strategy involves scaling its proven H-Optimus model—a widely used and best-in-class 1.1 billion parameter foundation model for histology images— into the comprehensive M-Optimus platform, capable of multimodal data integration and multiscale simulations. M-Optimus is fueled by a unique combination of large-scale multimodal public and proprietary data obtained through partnerships and large scale internal data generation. Its development is grounded in scientific rigor and adherence to strict regulatory privacy, and ethical data standards to ensure adoption across biopharma and healthcare. 

The opportunity:

AI to overcome inefficient drug R&D

Despite hundreds of billions of dollars invested each year in pharma R&D (Agrawal et al. 2025), drug discovery and development remains desperately slow, costly, and failure prone, taking 10 to 15 years, with costs averaging more than $2 billion US dollars to put a new therapy on the market, the cost being mainly driven by the low probability ~10% success rate for candidate drugs entering Phase I clinical trial achieves approval (DiMasi, Grabowski, and Hansen 2016; Philippidis 2023). This inefficiency creates a heavy economic and human burden, suppressing innovation and concentrating efforts on limited targets,

Please tell us your email address to continue reading.
We will email you a copy of the full manifesto, as well as unlocking the full text with chapters in your browser.