AI Materials Startup Deep Principle Closes Series A to Drive the Evolution of Materials R&D

HANGZHOU, China, October 1, 2026 (EZ Newswire) -- Deep Principle, a global leader in AI for materials, has completed its Series A round, bringing its cumulative funding to over $100 million. 

The company will use the proceeds to advance the AI Scientist platform, scale the in-house AI Materials Factory lab, accelerate the commercialization of key materials pipelines, and deepen collaboration with partners along the industry chain.

"We built strong models, data and our own lab, a closed loop that keeps refining the models, and a business that turns them into real materials for real customers," said Dr. Haojun Jia, CEO of Deep Principle.

Generative AI Models Keep Evolving Toward State of the Art

Deep Principle keeps pushing technical boundaries, releasing multiple models that achieve state-of-the-art (SOTA) results. In property prediction, the company launched the MPA model, which reached SOTA levels on 40 experimental property-prediction tasks. In materials generation, it developed SAGA, a model that can autonomously adapt within complex materials-generation tasks and demonstrates stronger multi-objective optimization than models developed by major global tech companies. In chemical reaction generation, the company released the OA-ReactDiff and React-OT models, opening up the black box of AI-driven predictions and shrinking transition-state generation from days or even months down to 0.4 seconds, making AI genuinely usable in industrial practice. 

Building on this full-stack SOTA model portfolio, the company built its AI scientist platform, Mira, which autonomously handles the design, execution, review, and dynamic optimization of end-to-end research workflows.

The Company's In-House Lab and Its Dry-Wet Closed Loop

Guided by "AI for Science, Science for AI," Deep Principle built AI Materials Factory, an L4 high-throughput lab — a self-driving lab where experiments run with minimal manual intervention — to bridge the "last mile" between AI prediction and experimental validation. Inside the lab, Mira designs experiments, operates equipment, and collects data on its own, forming a complete dry-wet closed loop: AI prediction in the 'dry' lab, high-throughput validation in the 'wet' lab, and results flowing back into knowledge accumulation and model iteration. The company focuses on computing-infrastructure materials, renewable energy, and superconductors, and will keep adding R&D pipelines this year to speed up the commercialization of frontier materials.

Repeat Business From Existing Clients, Expansion Into New Industries

Deep Principle has signed a number of leading clients around the world, including large multinational companies, many of which keep returning with repeat orders. The company's business models include providing platform services, co-developing material pipelines with partners, and developing its own pipelines in-house. It is now building proprietary pipelines in frontier areas such as liquid coolants for data centers. Its business now covers fine chemicals, nutrition and personal care, renewable energy, electronic materials, and other key new-materials sectors. In 2025, the Deep Principle secured orders totaling about US$1.4 million as it began scaling commercial deployments.

"Materials discovery has always relied on human experience and trial and error. AI is changing that. We are moving the field from experience-driven to intelligence-driven, and that shift is happening faster than most people realize," said Dr. Jia.

About Deep Principle

Founded in 2024 by Dr. Haojun Jia and Dr. Chenru Duan, both PhD graduates of MIT, Deep Principle focuses on applying AI to materials research and development. The company's mission is to "unlock breakthrough materials with AI." It combines generative AI with first-principles calculations to build a self-developed technology system spanning the entire materials R&D lifecycle, forming an end-to-end closed loop for materials development. Deep Principle is working to transform conventional experience-driven, trial-and-error-based materials chemistry R&D into an AI-powered precision design paradigm, aiming to compress the R&D cycle for new materials from 5-10 years to months. Its business spans renewable energy, electronic materials, and nutrition and personal care, with strategic positions in frontier areas such as superconductivity and data centers.

Media Contact

info@deepprinciple.com