COMPANY PROFILE / MATERIALS INFORMATICS
Schrödinger
— Materials software that adds machine learning to physics-based calculation, inside a business that is mostly drug discovery
Schrödinger is a US-listed company that sells software that calculates how molecules behave, using quantum mechanics and molecular dynamics, to pharmaceutical companies and materials makers. It has a materials product line (Materials Science Suite), develops machine-learning force fields that give the forces between atoms quickly and accurately, and in October 2026 announced a Nature paper on battery electrolytes.
For the company as a whole, though, drug discovery dominates. Reading the figure in the Form 10-K, annual contract value (ACV) from materials was about US$17.1 million in 2025, roughly 9% of the total (our calculation). This profile concentrates on the materials business and research, and separates the company’s performance claims from what has been checked by experiment.
- Schrödinger in 30 seconds, and where it stands in materials informatics
- The logo and how the name is written
- Where the company sits in the materials informatics field
- Scale (centered on materials)
- Resources committed to materials informatics
- What it has done so far
- What comes next
- What the company is aiming for in materials informatics
- Map of materials informatics partnerships
- Technology, products and services
- The risks this company carries
- Glossary
- References
- Claim-to-source audit
1. Schrödinger in 30 seconds, and where it stands in materials informatics
(primary)Secondary: machine-learning force fields (a foundation-model-like part)
Materials informatics (MI) is a way of finding new materials by collecting materials data, using AI or computation to predict and design properties, and then checking the result by experiment. Schrödinger sells software that does the “predict and design” step with physics-based calculation (quantum mechanics and molecular dynamics), and in recent years it has used machine learning to speed that calculation up Sourced. Unlike “data-driven” MI that learns statistically from data, calculation from the physical laws of atoms and molecules is its foundation.
But materials is a very small part of the business: about 9% of contract value, with most revenue and R&D going to drug discovery Our calculation. The view that materials science software is the center of the business does not hold on the official numbers.
2. The logo and how the name is written
A note on names: the company is written Schrödinger, Inc. (with an umlaut on the ö), but the registrant name in SEC filings is Schrodinger, Inc. (no umlaut). It is the same company. The name recalls the “Schrödinger equation” of quantum mechanics (we did not confirm the origin in official sources).
Official site: https://www.schrodinger.com/3. Where the company sits in the materials informatics field
Even when two companies both say “AI”, they differ in where they sit along the flow of collect data → predict and design → check by experiment. Here is Schrödinger placed on that flow.
For quantum calculation, see our explainers “Fundamentals of Density Functional Theory” and “Quantum Chemistry and Materials Simulation”; for molecular dynamics, “Molecular Dynamics”; and for machine-learning force fields, “Machine-Learning Interatomic Potentials”.
4. Scale (centered on materials)
Company-wide scale and the materials figures are shown separately Sourced.
| Item (US$ million) | FY2024 (to Dec 2024) | FY2025 (to Dec 2025) |
|---|---|---|
| Revenue (total) | 207.5 | 255.9 |
| of which software | 180.4 | 199.5 |
| of which drug discovery | 27.2 | 56.4 |
| R&D expense (total) | 201.8 | 173.1 |
| Operating loss | 209.3 | 166.9 |
| ACV (annual contract value, total) | 190.8 | 198.5 |
| Item | Detail |
|---|---|
| Employees (3 February 2026) | 850 full-time (363 with PhDs; 587 in the US and 263 outside) |
| Headquarters | New York, USA (1540 Broadway) |
| Listing | Nasdaq Global Select Market (SDGR), since 6 February 2020 |
| Segments | Two: software and drug discovery. Software includes both drug-discovery and materials science customers |
Materials figures (the ACV breakdown)
Starting in 2025 the company discloses ACV by industry. The materials figure can be read from a figure in the Form 10-K.
| ACV by industry (US$ million) | 2024 | 2025 | Share of the 2025 total (our calculation) |
|---|---|---|---|
| Materials | 14.9 | 17.1 | about 8.6% |
| Top 20 pharma | 70.0 | 80.8 | about 40.7% |
| Rest of life sciences | 92.4 | 90.4 | about 45.5% |
| Contribution | 13.5 | 10.2 | about 5.1% |
| Total | 190.8 | 198.5 | 100% |
Source: Reference 2, 1. The shares are our calculation Our calculation. Materials ACV grew about 15% from 2024 to 2025 (14.9 to 17.1) and the total about 4% (190.8 to 198.5).
The materials ACV is a figure in which customers who buy both drug-discovery and materials software are split by the company’s internal judgment (per the 10-K). ACV is not revenue and is not a forecast of revenue. “Contribution” is grant funding from Gates Ventures and the Bill & Melinda Gates Foundation, treated separately from software contracts. Materials revenue, R&D expense and profit are not disclosed Not yet confirmed.
5. Resources committed to materials informatics
Headcount and budget for materials are not disclosed, but the official sources show what is being put in. Items we could not confirm are marked as not found.
| Type of resource | What we found | Type |
|---|---|---|
| People | The company has 850 full-time employees (363 with PhDs). Dedicated materials headcount is not disclosed. The 10-K says it keeps building a team of subject-matter experts in each materials area. Mat Halls, senior vice president of materials science, appears in a 2026 news release | Sourced (headcount not disclosed) |
| Money | R&D expense was US$173.1 million company-wide in 2025. It combines drug-discovery program costs and platform development, and the materials share is not disclosed. As outside funding, the company signed a three-year agreement with Gates Ventures in June 2020 to develop atomistic simulation methods to improve battery performance, and extended it for three years at an increased scale in August 2023 | Sourced (materials breakdown not disclosed) |
| Equity stakes, acquisitions, joint ventures, divestitures | The 10-K contains no mention of equity stakes, acquisitions, joint ventures or sales related to materials | Not found |
| Universities, research institutes and public funding | A joint study with Columbia University and the New York Structural Biology Center (the Nature paper, October 2026); co-founder Richard Friesner is a professor at Columbia. A research collaboration with Copernic Catalysts on catalysts (2023). Public funding such as DOE, NSF or NEDO is not mentioned in the sources we checked | Sourced (public funding not found) |
Status of the main products and models
This is what the company’s official material lets us confirm. Performance figures are claims by the company or study team and were not reproduced or verified for this article.
| Name | Kind | Content (published specifications, target, conditions) | Release form and license | Stage | Date | Source |
|---|---|---|---|---|---|---|
| Materials Science Suite | A set of calculation software for materials | MS Maestro (interface), MS LiveDesign, Desmond (molecular dynamics), Jaguar (quantum mechanics), Formulation ML (automated machine learning for formulation-property relationships) and more. Applications include organic electronics, polymers, consumer goods, catalysis, thin films, batteries, and metals and ceramics | Commercial. Licenses by amount of calculation (per the 10-K) | Offered | 2026-3 is the latest (no date stated) | Reference 6, 7, 1 |
| Machine-learning force field (MPNICE) | Machine-learning interatomic potential | Predicts atomic charges iteratively and handles long-range interactions. The preprint reports inference 5 to 20 times faster than models of comparable accuracy; checks against experimental liquid and solid data for organic systems; and models for bulk inorganic crystals. It also reports predicting gas-phase structures of about 500 Pt and Ir organometallic complexes it was not trained on | The paper is public. How it is delivered in the products is not confirmed in official sources | Research (preprint) | May 2025 (v2 Aug 2025) | Reference 8 |
| Materials Science Suite 2026-3 | Latest release | Bunsen (beta), a custom extension of the base machine-learning force field, back-mapping from coarse-grained to all-atom models, machine-learning features for solids, and extended crystal-structure prediction | Commercial | Offered | No date stated | Reference 7 |
| Bunsen | Agentic AI (co-scientist) | Understands a scientific goal, plans a computational strategy, runs validated physics-based calculations and interprets results. NVIDIA and Google Cloud provide the computing platform | Commercial. The company describes a throughput-based licensing model | Early access (full version expected by the end of 2026) | Announced 27 Jul 2026 | Reference 5, 3 |
6. What it has done so far
| When | What happened | Why it matters |
|---|---|---|
| Feb 2020 | Listed on Nasdaq | As a company selling both drug-discovery and materials software |
| Jun 2020 | A three-year agreement with Gates Ventures on atomistic simulation to improve battery performance | The start of outside-funded materials research |
| 2023 | A research collaboration with Copernic Catalysts on sustainable catalysts for e-fuels and bulk chemicals | A collaboration in catalysis |
| Aug 2023 | The Gates Ventures agreement extended for three years at a larger scale | Battery research continues |
| May 2025 | The MPNICE (machine-learning force field) preprint released | Research on speeding up physics-based calculation with machine learning |
| FY2025 | The 10-K disclosed ACV by industry (materials US$17.1 million) | The scale of the materials business could be seen in numbers for the first time |
| 25 Feb 2026 | Form 10-K filed | Describes the materials business (application areas, Gates Ventures, Copernic) |
| 27 Jul 2026 | Early access to Bunsen announced | AI that drives physics-based calculation in natural language |
| 14 Oct 2026 | Announced a Nature paper on lithium-ion battery electrolytes (with Columbia University and others) | Results using machine-learning force fields. We have not checked the paper itself |
7. What comes next
end of 2026From the 10-Q; the company’s plan
The 10-Q says the company expects to release the full commercial version of Bunsen by the end of 2026 Not yet confirmed. This is the company’s plan, not an achieved fact.
None of the following is stated in the official sources we checked Not yet confirmed.
Materials revenue, profit and R&D expense; numerical targets for materials; how MPNICE and similar models are offered in products; and details of the Nature paper itself (this article relies on the company’s news release). Press-based talk that materials will grow by some multiple in some year is not used here.
8. What the company is aiming for in materials informatics
This part includes things the company does not say directly. It is our reading of the primary sources, written with the evidence for each point.
| What we read from it | Primary evidence | Type |
|---|---|---|
| The materials business is the drug-discovery physics platform extended sideways | The 10-K explains that the physics behind drug molecules and materials is the same, so the platform was extended to materials science | Sourced |
| The direction is speeding physics-based calculation with machine learning, not learning purely from data | The news release says the technology removes the trade-off between quantum precision and large-scale simulation; Bunsen “executes validated computational methods” | Inference |
| The company itself sees materials adoption as still early | The 10-K says the materials science industry is only beginning to recognize the potential of computational methods | Sourced |
| Battery research funded from outside (Gates Ventures) is pushing materials technology forward | The agreement running since 2020 and the 2026 Nature announcement | Inference |
| It aims to widen the user base for physics-based calculation with AI | The Bunsen release says it makes advanced methods accessible to researchers who are not computational chemists | Inference |
The 10-K says the same platform is used for both drug molecules and materials, and materials contract value (US$17.1 million) is about one tenth of the total for drug-discovery customers (top 20 pharma US$80.8 million plus the rest of life sciences US$90.4 million, US$171.2 million) Our calculation. The Nature paper and the Bunsen announcement, meanwhile, foreground technology that works for both materials and drug discovery. From this we read the company as spreading a shared calculation technology into one more field of use rather than growing materials as a separate business Not yet confirmed.
9. Map of materials informatics partnerships
10. Technology, products and services
Physics-based calculation, and the machine-learning force field
Methods for predicting a material’s properties by calculation trade accuracy against scale. Quantum-mechanical calculation (such as Jaguar) works from how electrons behave and is accurate, but handles few atoms and takes a long time. Molecular dynamics (such as Desmond) moves many atoms for long times using the forces between atoms (a force field), but the force field’s accuracy decides the result. A machine-learning force field learns from quantum-mechanical results to build a force field, aiming for “near-quantum accuracy at molecular-dynamics scale” (the company’s description).
According to the news release, the study showed a structure not recognized before: lithium ions surrounded by the solvent (ethylene carbonate) gather in groups without anions Sourced (the research team’s claim). Quantum-chemical calculations indicate these clusters are easily reduced and may be intermediates in forming the protective film on the anode (SEI) (a suggestion).
Three points here matter to a materials engineer Sourced. (1) How accuracy is checked: the release cites agreement with NMR spectra, density, viscosity, ionic conductivity and diffusion coefficients. For your own material, those measurements are the yardstick for judging a machine-learning force field. (2) Range of application: the co-founder’s comment says many complex chemical systems are now open to prediction, but that is an outlook, and what has been shown is the electrolyte example. It may not carry over directly to polymers, catalysts or solids.
(3) Quantum calculation underneath: a machine-learning force field learns from quantum data, so accuracy may fall outside the chemistry it was trained on (a general property, not individually confirmed in company material). The preprint’s “5 to 20 times faster” is against models of comparable accuracy, and the comparison targets and conditions need to be checked in the paper.
11. The risks this company carries
| Risk | Detail | Type |
|---|---|---|
| Materials is a very small part | About 9% of ACV. Growth and results depend heavily on drug-discovery contracts and the drug-discovery business | Our calculation |
| Materials adoption is early | The 10-K says the materials science industry is only beginning to recognize computational methods and there is no assurance of adoption | Sourced |
| Rivals exist | The 10-K names BIOVIA and Materials Design as rivals in materials science | Sourced |
| The breakdown rests on internal judgment | Materials ACV splits customers who buy both kinds of software by the company’s internal judgment | Sourced |
| Performance figures are company or study-team claims | The Nature paper’s “up to 40%” and MPNICE’s “5 to 20 times faster” have limited conditions, and we found no independent verification | Sourced |
| New product is at an early stage | Bunsen is in early access, with the full version “expected by the end of 2026” | Not yet confirmed |
| Overall profit and loss | Operating loss of US$166.9 million in 2025 (US$209.3 million in 2024) | Sourced |
| Reliance on outside grants | Battery research is supported by Gates Ventures. The October 2026 release says “ongoing”, but we could not confirm the terms after August 2026 | Inference |
12. Glossary
- Materials informatics (MI)
- Finding new materials with materials data plus AI and computation, in the flow of data, then predict and design, then experiment.
- Quantum-mechanical calculation
- Calculation of a molecule’s or material’s energy and properties from how its electrons behave. Accurate but heavy.
- Molecular dynamics
- A method that moves atoms according to the laws of force to simulate many atoms over long times.
- Force field
- The equations or model that give the forces between atoms from their positions.
- Machine-learning force field
- A machine-learning model trained on quantum calculations that gives forces between atoms quickly and accurately.
- Electrolyte
- The liquid that carries ions inside a battery. In lithium-ion cells it is a salt dissolved in organic solvents.
- SEI
- The protective film that forms on the anode surface. It affects battery life and safety.
- ACV
- Annual contract value. Not revenue.
- NMR
- Nuclear magnetic resonance, a measurement that probes the surroundings of molecules.
- Agentic AI
- AI that understands a goal, plans the steps and carries them out with tools.
- Throughput-based licensing
- Pricing by the amount of calculation processed.
- Form 10-K / 10-Q
- The annual and quarterly reports that US-listed companies file with the SEC.
13. References
- Schrödinger Form 10-K for the year ended 31 December 2025 (filed 25 February 2026). Business overview, materials science software, revenue, R&D expense, headcount and risk factors. The registrant name is “Schrodinger, Inc.”. https://www.sec.gov/Archives/edgar/data/1490978/000149097826000010/sdgr-20251231.htm
- Schrödinger Figure “ACV by Customer and Industry Cohort” in the Form 10-K above (2025 and 2024 ACV by industry and customer group). https://www.sec.gov/Archives/edgar/data/1490978/000149097826000010/sdgr-20251231_g2.jpg
- Schrödinger Form 10-Q for the quarter ended 30 June 2026 (filed 5 August 2026). Status of Bunsen; competitors in materials science. https://www.sec.gov/Archives/edgar/data/1490978/000149097826000068/sdgr-20260630.htm
- Schrödinger News release (14 October 2026) “Schrödinger Research Published in Nature Reveals New Insights into Lithium-Ion Battery Chemistry”. https://s203.q4cdn.com/609444515/files/doc_news/2026/10/14/Schrdinger-Research-Published-in-Nature-Reveals-New-Insights-into-Lithium-Ion-Battery-Chemistry.pdf
- Schrödinger News release (27 July 2026) “Schrödinger Introduces Bunsen, an AI Co-Scientist for Molecular Discovery”. https://s203.q4cdn.com/609444515/files/doc_news/Schrdinger-Introduces-Bunsen-an-AI-Co-Scientist-for-Molecular-Discovery-2026.pdf
- Schrödinger Official website, materials science page. Product names and application areas. https://www.schrodinger.com/materials-science
- Schrödinger Official website, release notes. Features of Materials Science Suite 2026-3. https://www.schrodinger.com/release-notes
- Weber et al. “Efficient Long-Range Machine Learning Force Fields for Liquid and Materials Properties”, arXiv:2505.06462 (May 2025; v2 August 2025). MPNICE. https://arxiv.org/abs/2505.06462
- Schrödinger Official website (source of the logo; used to check the company profile). https://www.schrodinger.com/
14. Claim-to-source audit
| Claim in the text | Type | Source |
|---|---|---|
| Legal name, headquarters, listing (since 6 February 2020) and 850 employees | Sourced | Reference 1 https://www.sec.gov/Archives/edgar/data/1490978/000149097826000010/sdgr-20251231.htm |
| Revenue, software and drug-discovery split, R&D expense and operating loss (2024 and 2025) | Sourced | Reference 1 https://www.sec.gov/Archives/edgar/data/1490978/000149097826000010/sdgr-20251231.htm |
| The ACV breakdown (materials 17.1, top 20 pharma 80.8, rest of life sciences 90.4, contribution 10.2, total 198.5 in US$ million; 2024: 14.9, 70.0, 92.4, 13.5, 190.8) | Sourced | Reference 2 https://www.sec.gov/Archives/edgar/data/1490978/000149097826000010/sdgr-20251231_g2.jpg |
| Materials share of about 8.6% (2025) and about 7.8% (2024), and the growth comparison | Our calculation | Our own calculation (this article), from the figures in Reference 2 |
| Application areas in materials science, the Gates Ventures agreement (June 2020, August 2023) and the Copernic Catalysts collaboration | Sourced | Reference 1 https://www.sec.gov/Archives/edgar/data/1490978/000149097826000010/sdgr-20251231.htm |
| Rivals in materials science (BIOVIA, Materials Design) and the early stage of materials adoption | Sourced | Reference 1 https://www.sec.gov/Archives/edgar/data/1490978/000149097826000010/sdgr-20251231.htm Reference 3 https://www.sec.gov/Archives/edgar/data/1490978/000149097826000068/sdgr-20260630.htm |
| Product names and application areas of Materials Science Suite | Sourced | Reference 6 https://www.schrodinger.com/materials-science |
| Features of Materials Science Suite 2026-3 (Bunsen beta, force-field extension and more) | Sourced | Reference 7 https://www.schrodinger.com/release-notes |
| MPNICE content and performance claims (preprint) | Sourced | Reference 8 https://arxiv.org/abs/2505.06462 |
| The Bunsen announcement (27 July 2026), the NVIDIA and Google Cloud collaboration, and the throughput-based licensing description | Sourced | Reference 5 https://s203.q4cdn.com/609444515/files/doc_news/Schrdinger-Introduces-Bunsen-an-AI-Co-Scientist-for-Molecular-Discovery-2026.pdf |
| The statement that the full commercial version of Bunsen is expected by the end of 2026 | Not yet confirmed | Reference 3 https://www.sec.gov/Archives/edgar/data/1490978/000149097826000068/sdgr-20260630.htm |
| The announced Nature paper on electrolytes (14 October 2026), lithium-ion clusters, the experimental comparison items and the Gates Ventures funding (company claims) | Sourced | Reference 4 https://s203.q4cdn.com/609444515/files/doc_news/2026/10/14/Schrdinger-Research-Published-in-Nature-Reveals-New-Insights-into-Lithium-Ion-Battery-Chemistry.pdf |
| Placement on the three steps of MI (data, predict and design, experiment) | Our calculation | Our own grouping from public sources (this article) |
| Materials revenue, profit, R&D expense and dedicated headcount | Not yet confirmed | None could be confirmed, and this article does not estimate them (not stated in the sources checked for this article) |
| Equity stakes, acquisitions, joint ventures and public funding (DOE, NSF, NEDO and so on) for materials | Not yet confirmed | Not stated in Reference 1 as checked for this article; we do not assert that none exist |
| Details of the Nature paper, how MPNICE is offered in products, and numerical targets for materials | Not yet confirmed | Could not be confirmed; this article does not guess |
| Reading: the direction is speeding physics-based calculation with machine learning | Inference | Our own interpretation of the news release and the Bunsen description |
| Reading: materials is spread as one more field of use of a shared technology | Inference | Our own interpretation of the 10-K and the 2026 announcements |
| Reading: battery research is pushing materials technology forward | Inference | Our own interpretation of the Gates Ventures agreement and the announcement |
| Reading: the aim is to widen the user base for calculation | Inference | Our own interpretation of the Bunsen description |
| Risk: reliance on outside grants | Inference | Our own interpretation of the agreement terms |
Last updated 22 October 2026 / Troy Technical
The figures in this article rest on Schrödinger’s SEC filings (Form 10-K and 10-Q), official news releases, the official website and a preprint. Market-research estimates and press-only figures are not used. Figures described as claims by the company or study team were not reproduced or verified here. Items marked “Inference” are our own reading of the primary sources and are not statements by the company.