Executive Summary / Key Takeaways

• SES AI is undergoing a strategic transformation, pivoting to an AI-first, asset-light business model that leverages its Molecular Universe for material discovery and Avatar for battery safety across diverse markets.
• This shift is driving significant revenue growth and impressive gross margins (74-79% in H1 2025), primarily from high-value software and service offerings, positioning the company for scalable expansion.
• Technological leadership in AI-enhanced Li-Metal and high-silicon Li-ion cells, evidenced by Li-Metal B-samples passing rigorous EV safety tests and UAM cells offering superior energy density, underpins its competitive differentiation.
• The recent acquisition of UZ Energy is a pivotal move, accelerating SES AI's entry into the rapidly growing Battery Energy Storage Systems (ESS) market, which is projected to be the primary revenue driver for 2025.
• With a strong liquidity position of $229 million (Q2 2025) and a CapEx-light approach, SES AI is well-funded to execute its strategy, though external risks like changing EV incentives and trade policies warrant close monitoring.

The AI-Powered Energy Transition: SES AI's Strategic Reimagining

SES AI Corporation, founded in 2012, has embarked on a profound strategic evolution, moving from a traditional, capital-intensive battery developer to an AI-first powerhouse aiming to accelerate the world's energy transition. Initially focused on high-performance Lithium-Metal (Li-Metal) batteries for electric vehicles (EVs), the company's journey has been marked by significant joint development agreements (JDAs) with leading automotive OEMs like Hyundai (HYMTF) and Honda (HMC). This foundational work in advanced battery chemistry laid the groundwork for a transformative "All-in on AI" strategy, which now defines its competitive edge in a rapidly expanding global market.

The broader industry landscape is characterized by an escalating demand for advanced energy solutions, driven by the proliferation of EVs, the burgeoning Urban Air Mobility (UAM) and drone sectors, and the critical need for robust Battery Energy Storage Systems (ESS) to power the AI era. In this dynamic environment, SES AI positions itself as an emerging innovator, leveraging its unique AI capabilities to address fundamental challenges in battery material discovery and safety. While direct quantitative market share figures for all niche competitors are not publicly detailed, the company's strong performance in its specialized segments suggests a growing influence. Its primary direct competitors, such as QuantumScape (QS), Solid Power (SLDP), and Enovix (ENVX), are also in R&D-heavy phases, often grappling with negative profitability and cash flow as they pursue their own advanced battery technologies. SES AI's strategic response to this competitive intensity is to differentiate through unparalleled speed of innovation and a capital-light business model.

Technological Edge: Molecular Universe and Avatar Platforms

At the heart of SES AI's transformation are its two proprietary AI platforms: Molecular Universe for upstream material discovery and Avatar for downstream battery safety. These platforms represent a significant technological moat, offering capabilities that management believes are unmatched in the industry.

Molecular Universe (MU) is a powerful Large Language Model (LLM) specifically trained on the world's most comprehensive battery domain expertise. It integrates the largest and most accurate battery-relevant molecule property databases, computed through GPU-accelerated large-scale computational chemistry algorithms. This platform's "Deep Space" feature, introduced in Molecular Universe 0.5, boasts an agentic capability that can conduct senior scientist-level battery research, aiming to reduce product development time from years to mere tens of minutes. The company has already mapped 10^8 (0.1%) of the molecular universe, which it asserts is the world's largest single-molecule density function theory database. This capability allows SES AI to accelerate novel electrolyte discovery, expecting to identify a new electrolyte every six months, a significant acceleration from the previous 9-12 months per discovery. For investors, this translates into a sustainable competitive advantage, enabling rapid product iteration and the potential for high-margin software and service revenue streams.

Complementing MU is the Avatar platform, a new foundation model designed to monitor battery health and predict incidents before they occur. Avatar works in conjunction with conventional physics-based Battery Management Systems (BMS) and is uniquely agnostic to battery chemistry, form factor, and mission profile. It has been pretrained on manufacturing and cycling data across a wide variety of cell chemistries, including NCM, LFP, high-silicon graphite, and Li-Metal anodes, as well as pouch, prismatic, and cylindrical form factors. The platform has demonstrated 95% prediction accuracy with AI for Safety alone, achieving 100% accuracy when integrated with AI for Manufacturing, based on training from 15,000 lithium metal cells. This robust safety capability is critical for commercialization, particularly in high-stakes applications like EVs and UAM, and enhances the value proposition of SES AI's battery products.

SES AI's technological prowess extends to its battery cells themselves. Its 100 mPOWER Li-Metal B-sample cells have successfully passed the rigorous GB38031-2020 EV safety standard, an industry first for large-capacity Li-Metal cells. For UAM applications, SES AI's Li-Metal cells offer 400 watts per kilogram at the cell level, significantly outperforming conventional Li-ion power cells at approximately 250 watts per kilogram. This translates to a tangible benefit of enabling 50% longer flight times or 50% more payload capacity, a critical differentiator in the rapidly growing drone and UAM markets. Furthermore, the new electrolytes discovered through Molecular Universe enable stable performance and prevent gassing for 2170 cells with high silicon content, achieving over 6.5 mPOWERs and even 7 mPOWERs, and are expected to more than double the cycle life for 100% silicon anodes. This addresses a key challenge for high-silicon Li-ion batteries, which typically suffer from swelling and short cycle life, opening new market opportunities for SES AI in robotics and power tools.

Strategic Evolution and Financial Performance

SES AI's strategic shift is evident in its business model evolution from CapEx-heavy manufacturing to an asset-light approach focused on selling AI models, core battery materials, and utilizing contract manufacturing for cells. This pivot has enabled earlier revenue generation and a more scalable business. The company now boasts diversified revenue streams encompassing software and service subscriptions, sales of novel electrolytes and battery materials, sales of AI-enhanced Li-Metal and high-silicon Li-ion cells, EV development services, and integrated ESS hardware/software solutions.

This strategic reorientation began to bear fruit in October 2024, when SES AI first started deriving revenue from its principal business activities. The company reported its inaugural revenue of $2.0 million in Q4 2024, followed by $5.8 million in Q1 2025 and $3.5 million in Q2 2025, totaling $9.3 million for the first half of 2025. This revenue was primarily driven by service-related contracts with automotive OEMs for AI-enhanced battery materials. Crucially, these revenue streams have delivered impressive gross margins: 63% in Q4 2024, 79% in Q1 2025, and 74% in Q2 2025. Management attributes these high margins to the asset-light model and the significant contribution from high-margin software and service components, which are expected to yield over 80% margins, compared to product sales at 20-30%.

A pivotal strategic move occurred on July 25, 2025, with the agreement to acquire Shenzhen UZ Energy Co., Ltd., a China-based ESS manufacturer, for approximately $25.5 million. This acquisition, expected to close in Q3 2025, provides SES AI a critical foothold in the fast-growing global ESS market. UZ Energy's projected revenue for FY2025 is $10 million to $15 million, with expectations for "exponentially larger growth" in subsequent years. This integration will combine UZ's ESS hardware with Molecular Universe's material development capabilities for LFP and sodium-ion chemistries, and Avatar's precise health monitoring, positioning ESS as the "#1 use for Molecular Universe" and the bulk of SES AI's 2025 revenue.

While the company is now revenue-generating, it remains in an early growth stage, incurring net losses of $22.65 million in Q2 2025 and $35.08 million for H1 2025, consistent with its heavy investment in R&D. R&D expenses for Q2 2025 increased by 27% to $19.1 million (48% to $39.6 million for H1 2025), primarily due to a substantial increase in AI infrastructure costs, including GPU expenses and Molecular Universe development. This was partially offset by reductions in personnel and stock-based compensation, reflecting a focused allocation of resources. General and administrative expenses also saw a significant decrease, down 32% in Q2 2025 to $6.5 million, largely due to headcount reductions and lower professional services.

SES AI maintains a strong liquidity position, with $229 million in cash and marketable securities as of June 30, 2025, and no debt. Cash used in operations significantly decreased by 51% year-over-year and 53% quarter-over-quarter to $10.8 million in Q2 2025, demonstrating management's operational discipline. Capital expenditures were a modest $1.7 million for H1 2025, a substantial decrease from the prior year, as the company shifts investment towards AI-related infrastructure. The company's Board authorized a $30 million share repurchase program in April 2025, and in Q3 2025, SES AI repurchased and canceled 871,754 shares for $1.1 million, signaling management's confidence in its valuation and financial health.

Competitive Landscape: Differentiating in a Dynamic Market

SES AI operates in a highly competitive and evolving battery market, where its AI-first strategy and technological differentiators are crucial for carving out market share against both emerging and established players. Direct competitors like QuantumScape, Solid Power, and Enovix are similarly focused on next-generation battery technologies, primarily for EVs. These companies, like SES AI, are in R&D-intensive phases, characterized by negative profitability and cash flow as they scale their technologies.

Compared to QuantumScape, which focuses on solid-state batteries for safety and charging speed, SES AI's lithium-metal technology offers potentially higher energy density, which could translate to superior range and performance in EVs. While QuantumScape has established visible partnerships, SES AI aims for faster market entry through its asset-light model and rapid innovation cycles. Against Solid Power, another solid-state battery developer, SES AI's lithium-metal cells offer a significant advantage in energy density, potentially leading to better performance in high-demand EV applications. Solid Power's strength lies in its safety-focused technology and collaborative market entry, while SES AI emphasizes performance and faster charging capabilities. Enovix, specializing in silicon-anode batteries for consumer electronics and EVs, focuses on high-capacity designs and manufacturing innovation. SES AI's new electrolytes for high-silicon Li-ion cells offer materially improved cycle life and address swelling issues, potentially providing a lower operating cost advantage in certain applications compared to Enovix's offerings.

SES AI's primary competitive moat lies in its unique AI platforms. The Molecular Universe's ability to map the molecular space and accelerate electrolyte discovery, coupled with Avatar's precise battery health monitoring, provides an unparalleled speed of innovation and access to proprietary data. This allows SES AI to develop solutions that management believes are "much better than the best human scientists can." While the company acknowledges that successful new materials can be copied, its focus on the "speed of innovation" and selling the "map" and "tool" for discovery, rather than just individual materials, creates a more defensible position. However, SES AI faces competitive disadvantages in scaling challenges and high R&D dependencies, which could impact its financial performance and market entry timelines. Indirect competitors include traditional lithium-ion manufacturers and alternative energy storage solutions, which could offer more accessible or easier-to-implement options in certain non-EV applications.

Outlook and Risks: Charting the Path Forward

SES AI has affirmed its full-year 2025 revenue guidance of $15 million to $25 million, a significant milestone reflecting its transition to a commercialized entity. Management forecasts total cash usage in operations and capital expenditures for 2025 to be between $70 million and $80 million, prioritizing R&D advancements, commercialization, and strategic market expansion. The company expects to exit 2025 with over $200 million in liquidity, projecting its current financial runway to extend well into 2028, supported by its CapEx-light business model. The bulk of 2025 revenue is anticipated to come from the newly acquired ESS segment, with the remainder from EV and drones/robotics, signaling a successful diversification strategy.

However, the path forward is not without risks. Regulatory and economic incentives play a crucial role in the EV and energy storage markets. The recently enacted One Big Beautiful Bill Act (OBBBA) in July 2025, which terminates tax credits for EV purchases and charging infrastructure, could reduce EV demand and potentially delay battery product deliveries. Geopolitical tensions and changes in U.S. and foreign trade policies, including tariffs on imports from China (a key source of battery supplies), could increase costs and impact gross margins. The company also faces the ongoing risk of stock price volatility, as evidenced by its recent NYSE delisting notice (though compliance was regained), which could affect its market perception and access to capital. Internally, a material weakness in internal control over financial reporting related to Sponsor Earn-Out liabilities requires ongoing remediation. While current liquidity is strong, future funding may be required for additional acquisitions, AI model development, and unexpected delays in commercialization.

Conclusion

SES AI Corporation stands at a pivotal juncture, transforming from a pioneering Li-Metal battery developer into a dynamic, AI-first platform company. Its "All-in on AI" strategy, powered by the Molecular Universe for accelerated material discovery and Avatar for enhanced battery safety, represents a compelling investment thesis. This strategic pivot has enabled an asset-light business model, driving impressive gross margins and diversified revenue streams across the high-growth EV, UAM, and ESS markets. The acquisition of UZ Energy is a testament to this strategic agility, providing an immediate foothold in the burgeoning ESS sector and a clear path to integrating physical AI solutions.

While the company is still in its early commercialization phase, evidenced by ongoing net losses, its strong liquidity, disciplined capital management, and affirmed revenue guidance for 2025 underscore a robust financial foundation. The technological leadership in AI-enhanced battery materials and safety offers a distinct competitive edge, positioning SES AI to address critical performance and safety needs in a rapidly evolving energy landscape. Investors should recognize SES AI as a high-potential play on the future of energy storage, where its unique AI capabilities and strategic market expansion could unlock significant long-term value, despite the inherent risks associated with a disruptive technology company.