WOBURN, Mass.--(BUSINESS WIRE)--May 13, 2026--
Molecular Universe Pte Ltd (“Molecular Universe”) today announced the release of Molecular Universe MU-3.0, its latest AI material discovery platform. MU-3.0 introduces MU-StarSeeker, an agent-managed, AI-enabled material discovery workflow automation capable of both cloud and on-premises deployment and purposely built to accelerate scientific research and product development across drones, robotics, energy storage, and mobility.
This press release features multimedia. View the full release here: https://www.businesswire.com/news/home/20260513965558/en/
Traditional battery R&D is slow, fragmented, and resource-intensive. Developing and commercializing a new chemistry could take 10 years to go from lab-scale R&D through A-sample, B-sample, C-sample, and finally SOP, involving tedious manual iterations across materials search, formulation, simulations, testing, and validation. Starting with MU-3.0 release, MU-StarSeeker is built to disrupt that traditional model by turning isolated research steps into automated and continuously learning agent-managed workflow.
“MU-3.0 is the most exciting Molecular Universe release since our original release of MU-0 a year ago,” said Qichao Hu, founder and CEO at SES AI Corporation, the parent company of Molecular Universe. “Starting with MU-3.0, everything now works for sodium just as well as it does for lithium chemistries, and MU-StarSeeker agent with its dry and wet data integrated loops can help users manage complex material discovery workflows. Users can literally set it and forget about it and come back and be happily surprised by the discoveries.”
Molecular Universe team will host a live webinar at 11am EDT, May 27 th, 2026. Please register at https://www.ses.ai/mu3-demo.
Key Advances in MU-3.0 include:
1.MU-StarSeeker, agent-managed material discovery workflow automation
Starting with MU-3.0, MU-StarSeeker transforms Molecular Universe’s core workflow features— Ask, Search, Formulate, Design, Predict, and Manufacture —into a Skill-based automation architecture. Users can combine Skills into customized workflow and delegate complex, multi-step workflows to MU-StarSeeker agent that orchestrates the relevant Skills end-to-end. For example, instead of manually searching molecules, testing formulations, evaluating bulk properties, validating cell-level performance, and repeating the process across hundreds of molecules and thousands of formulations, users can launch MU-StarSeeker agent workflow before heading on vacation—and return to a ranked list of 20+ top formulations optimized for their desired cell-level performance based on the feedback from all relevant features.
MU-StarSeeker includes standard Skills created and maintained by Molecular Universe and can work with individual and enterprise users to create and maintain their own Skills to capture proprietary workflows, best practices and domain expertise.
MU-StarSeeker agent is a major leap forward in workflow automation and standardization across organizations. At the same time, the six individual features will continue to be available to users who prefer more transparency and in-depth hands-on involvement.
2.MU-3.0 supports both lithium and sodium chemistries
With their low cost, high safety, superior fast charging and low temperature performance, and abundant and sustainable nature as compared with Lithium-ion (Li-ion) batteries, Na-ion batteries are quickly gaining popularity and market shares in energy storage applications. Na-ion chemistry is also one of the most searched prompts in Molecular Universe. Starting with MU-3.0, all services have been expanded to cover Na-ion. The popular molecular dynamics simulation engine and our proprietary force field in Formulate has been reengineered from the ground up to work just as well for Na-ion as it did for Li-ion, providing robust and accurate property prediction and analysis in a timely manner.
3.Closed-loop dry and wet data through autonomous labs or A-labs
MU-3.0 connects computational prediction with experimental validation through integrated dry and wet data workflows. Users can send wet data requests through individual feature access or MU-StarSeeker agent-managed Skill automation to MU’s high-throughput autonomous labs or A-labs.
This creates a closed-loop discovery workflow: MU-StarSeeker agent generates hypotheses, simulations screen molecule and formulation candidates, and predicts cell level performance, and A-labs produce experimentally ground-truth data, and the platform learns from the results. This approach can reduce unnecessary experiments, improve prediction accuracy, and accelerate the discovery of safer, higher-performing chemistries. Molecular Universe’s most important assets are its ever-growing high-quality high-fidelity dry and wet data. Simulated prediction results (dry data) can never be as accurate as experimentally measured data (wet data), and there are certain properties whose computation models don’t even exist yet, while it’s just easier to experimentally measure them.
For sensitive user data, MU can also provide dry and wet data on-premises using different forms such as Search-in-a-Lab or Formulate-in-a-Lab where the MU-3.0 features or Skills are incorporated with a high-performance computer and an A-lab and deployed locally at the user site.
4.Enterprise deployment and business impact
MU-3.0 is available through both cloud and on-premises deployment, supporting enterprise requirements for scalability, speed, security, and data control.
For large enterprise users, MU offers “in-a-box” solutions including Search-in-a-Box, Formulate-in-a-Box, and Predict-in-a-Box. These offerings support single-molecule property search, formulation-level bulk property simulations, and device-level performance predictions, and are already generating revenue and bookings from large enterprise users.
By compressing R&D timelines, reducing experimental wastes, standardizing best practices, and accelerating the path from chemistry concept to product deployment, MU-3.0 creates significant value for customers in drones, robotics, energy storage, and mobility.
Molecular Universe is a wholly-owned subsidiary of SES AI Corporation.
Forward-Looking Statements
This press release contains forward-looking statements within the meaning of Section 27A of the Securities Act of 1933, as amended, and Section 21E of the Securities Exchange Act of 1934, as amended, about SES AI and Molecular Universe that involve substantial risks and uncertainties. Forward-looking statements generally relate to future events or our future financial or operating performance. In some cases, you can identify forward-looking statements because they contain words such as “will,” “goal,” “prioritize,” “plan,” “target,” “expect,” “focus,” “look forward,” “opportunity,” “believe,” “estimate,” “continue,” “anticipate,” “project” and “pursue” or the negative of these terms or similar expressions. These statements are based on the beliefs and assumptions of the management of SES AI and Molecular Universe. You should not place undue reliance on these forward-looking statements. Although SES AI believes that the plans, intentions and expectations reflected in or suggested by these forward-looking statements are reasonable, it cannot provide assurance that it will achieve or realize these plans, intentions or expectations. Should one or more of a number of known and unknown risks and uncertainties materialize, or should any of our assumptions prove incorrect, our actual results or performance may be materially different from those expressed or implied by these forward-looking statements. Some factors that could cause actual results to differ include, but are not limited to, among other things, the risk that the market for the Molecular Universe platform, and for the MU-StarSeeker and A-lab capabilities, is still emerging, and may not achieve the customer interest or growth potential that SES AI expects; potential supply chain difficulties; the ability to obtain raw materials, components or equipment through new or existing supply relationships; our use of artificial intelligence and machine learning may result in legal and regulatory risk; risks related to intellectual property; business, regulatory, political, operational, financial and economic risks related to business operations outside the United States; and other factors described in our filings with the Securities and Exchange Commission (the “SEC”), including in the “Risk Factors” and “Management’s Discussion and Analysis of Financial Condition and Results of Operations” sections of our most recently filed Annual Report on Form 10-K, Quarterly Report on Form 10-Q and other documents that we have filed, or that we will file, with the SEC. Any forward-looking statements made by us in this press release speak only as of the date on which they are made, and subsequent events may cause these expectations to change. We disclaim any obligations to update or alter these forward-looking statements in the future, whether as a result of new information, future events or otherwise, except as required by law.
Molecular Universe Introduces MU-StarSeeker, an Agent-Managed AI Platform as Part of its Latest Frontier Workflow Automation MU-3.0 Release.
