Periodic Labs, an artificial intelligence startup focused on materials science, has secured $200 million in funding at a $1 billion pre-money valuation, with Andreessen Horowitz leading the investment round. The company, founded by former OpenAI executives, represents the latest high-profile venture from the growing pool of talent departing the ChatGPT maker.
The financing underscores investor appetite for AI applications beyond chatbots and language models, particularly in scientific research where machine learning could accelerate discovery timelines. Sources close to the deal indicate that while OpenAI was initially expected to lead the funding, founders ultimately selected A16z for its broader strategic capabilities and resource network.
Leadership with Deep AI Expertise
The startup was established by Liam Fedus, OpenAI’s former vice president of research who played a central role in developing ChatGPT, and Ekin Dogus Cubuk, a research scientist previously with Google DeepMind. Their combined experience spans foundational work in large language models and machine learning applications for scientific problems.
Liam Fedus had previously expressed interest in applying AI to scientific discovery, making materials science a natural focus area. The company aims to develop computational tools that can model and predict material properties more accurately and rapidly than conventional research methods, potentially transforming development cycles across energy storage, electronics, and construction sectors.
The OpenAI Alumni Phenomenon
Periodic Labs joins a growing roster of ventures launched by former OpenAI personnel, a trend drawing comparisons to previous waves of entrepreneurship from Google and PayPal alumni. Anthropic, founded by siblings Dario and Daniela Amodei, exemplifies this pattern, as does Thinking Machines Lab, recently valued at $10 billion under former OpenAI executive Mira Murati’s leadership.
Investment data reveals that ventures led by ex-OpenAI employees have collectively attracted more than $42 billion in funding, reflecting investor confidence in their technical capabilities and market insight. Applied Compute, founded by former OpenAI researchers Rhythm Garg, Linden Li, and Yash Patil, recently achieved a $100 million valuation from Benchmark despite being in early development stages.
AI Dominance in Venture Capital
The Periodic Labs funding occurs during a period of strong investor interest in artificial intelligence startups. During the second quarter of 2025, eleven AI-focused companies raised over $1 billion combined, representing more than one-third of the quarter’s $91 billion global venture capital total.
July proved particularly active for AI investments, capturing 37% of all venture funding and exceeding healthcare ($5.7 billion) and financial services ($4.6 billion) sectors. Notable July transactions included xAI’s $5 billion raise led by SpaceX, marking the year’s third-largest venture deal behind OpenAI’s $40 billion SoftBank-led round in March and Meta’s $14.3 billion investment in Scale AI during June.
Materials Science Applications
By leveraging machine learning models to simulate and predict material properties, Periodic Labs seeks to compress research timelines that traditionally require years of experimental work. This approach could influence industries ranging from clean energy to semiconductors, potentially reshaping supply chains and technological development processes.
The A16z partnership provides Periodic Labs with capital and strategic guidance for commercialization and scaling efforts. Success in early projects will likely draw attention from both venture capital and scientific research communities, given the significant resources committed to the company’s development.
The deal terms remain subject to change as the financing has not yet closed, though the preliminary valuation reflects market confidence in AI-driven scientific applications. Periodic Labs’ rapid emergence highlights both competitive dynamics for AI talent and investor willingness to support science-focused applications at early stages.