Schrodinger

Schrödinger's case rests on a simple claim: that accurate physics plus curated force fields and machine-learned corrections produce numbers you can act on before you make the compound. You see it in the product line — FEP+ for binding affinity ranking, Desmond for molecular dynamics at scale, Jaguar for quantum mechanics, Glide for docking and virtual screening — all hosted inside Maestro so a medicinal chemist and a computational chemist can argue over the same structure. LiveDesign adds collaborative project tracking so those numbers live somewhere shared rather than in one person's notebook. The breadth is the other differentiator. Most competitors pick a lane: small-molecule docking, or MD, or biologics, or materials. Schrödinger covers all of them in one license framework. BioLuminate handles antibody design, peptide discovery, enzyme engineering, bifunctional degrader design and TCR modeling. MS Maestro handles polymers, organic electronics, energy capture and storage, thin film processing, catalysis, consumer packaged goods, and pharmaceutical formulations. If your organization spans drug discovery and materials science, having one vendor and one data model is worth real money in integration and training. Release 2026-3 shows where the company thinks the next fight is. Bunsen, the chemistry-native agentic co-scientist now in beta, is an attempt to put an assistant on top of the toolchain rather than bolting a generic chatbot onto it — it is built directly on the Schrödinger platform, which matters because an agent that can call FEP+ and Desmond is a different product from one that writes summaries. Macrocycle solutions in open beta go after cyclic peptides, an area where force-field approximations have historically struggled. Crystal Structure Prediction reaching full release with salts, solvates, and co-crystals is the quieter but more consequential item: solid-state form selection is where pharmaceutical development programs lose time and money, and having that in the same platform as the discovery work is a genuine workflow win. The Predictive Toxicology expansion to GPCR, bromodomain, and hormone receptors extends off-target risk screening earlier in the funnel. E-sol for Kp,uu helps you recognize compounds likely to penetrate the CNS. Where it strains. Everything physics-based here is compute-hungry — FEP+, Desmond, and Jaguar all want substantial hardware or cloud capacity at project scale, and that infrastructure cost sits on top of licensing. New users face a steep learning curve before their results are trustworthy, which is precisely why Schrödinger sells modeling services and certification courses alongside the software; the course catalog (Introduction to molecular modeling in drug discovery, free energy calculations with FEP+, computational antibody engineering) is a tell about how much training the product assumes. Bunsen is beta and Macrocycle solutions are open beta — treat both as evolving, not settled production