Loading organizations...
§ Private Profile · San Francisco, CA, USA
Periodic Labs is a technology company.
Periodic Labs develops artificial intelligence scientists engineered to accelerate discovery across the physical sciences. The company integrates advanced AI models with autonomous laboratory systems, automating the full scientific method. This approach focuses on generating hypotheses, conducting experiments, and refining understanding, streamlining the development of novel materials and phenomena.
The company was founded by Liam Fedus, a key contributor to large language models at OpenAI, and Ekin Dogus Cubuk, a distinguished AI researcher from Google Brain. Their insight recognized the potential to apply cutting-edge AI to scientific inquiry, automating the iterative cycle of conjecture and experimentation. They observed that traditional discovery could benefit from AI's ability to process vast datasets.
Periodic Labs targets sectors and institutions requiring breakthroughs in materials science, physics, and chemistry. Its long-term vision centers on enabling transformative scientific advancements, including the discovery of room-temperature superconductors, improved transportation, and more efficient energy systems. The company aims to drive fundamental shifts in scientific innovation.
Periodic Labs has raised $300.0M across 1 funding round.
Periodic Labs has raised $300.0M in total across 1 funding round.
Periodic Labs has raised $300.0M in total across 1 funding round.
Periodic Labs's investors include Peter Deng, Conviction Partners, Founders Fund, Innovation Endeavors, Kima Ventures, Paradigm, Pear VC, Stellar Capital, bill weathersby, Elad Gil, Florian Douetteau, John Collison.
Periodic Labs has raised $300.0M across 1 funding round. Most recently, it raised $300.0M Seed in September 2025.
| Date | Round | Lead Investors | Other Investors | Status |
|---|---|---|---|---|
| Sep 1, 2025 | $300M Seed | Peter Deng | Conviction Partners, Founders Fund, Innovation Endeavors, Kima Ventures, Paradigm, Pear VC, Stellar Capital, Bill Weathersby, Elad GIL, Florian Douetteau, John Collison, Karim Atiyeh, MEI Z., Tony XU, Eric Schmidt, Jeff Bezos, Jeff Dean, Vinod Khosla, Accel, Andreessen Horowitz, DST Global, NVIDIA, Brian Zhan, Striker Venture Partners | Announced |
Periodic Labs is an AI research company building autonomous laboratories powered by AI scientists to accelerate scientific discovery in materials science, chemistry, and physical experimentation.[1][2][3] It develops full-stack systems where AI agents collaborate with robotic labs to form hypotheses, run experiments, analyze data (often gigabytes per experiment), and iterate rapidly, generating unique datasets including negative results.[1][3][4] The company serves industries like semiconductors (e.g., improving chip heat dissipation), space, defense, energy, hardware, pharmaceuticals, and nuclear fusion, solving the bottleneck of slow, manual scientific R&D by enabling thousands of experiments daily.[2][3] With a 20-200 person team of elite talent from OpenAI, DeepMind, Google Brain, and Microsoft, and $300M in seed funding, Periodic Labs shows explosive early momentum through industry partnerships and a focus on verifiable physical sciences.[1][2][6]
Periodic Labs was founded in 2025 by Liam Fedus (co-creator of ChatGPT, former OpenAI VP of Research) and Ekin Dogus Cubuk (former DeepMind lead in materials science and chemistry).[1][2][3] The idea emerged from their frustration with internet-trained AI's limits in science—finite data can't match real-world experimentation—so they merged AI with physical labs for grounded discovery.[2][3] The founding team, around 20 strong, includes veterans behind DeepMind’s GNoME (materials discovery), OpenAI’s Operator agent, Microsoft’s MatterGen LLM, and neural attention mechanisms, plus experience scaling autonomous physics labs.[2][3] Launching September 30, 2025, in San Francisco, they immediately raised a massive $300M seed round led by Andreessen Horowitz (a16z), with Nvidia Ventures, Felicis, Accel, DST, and angels like Jeff Bezos, Eric Schmidt, and Jeff Dean—securing elite backing and early semiconductor partnerships as pivotal traction.[2][3][6]
Periodic Labs stands out in AI-driven science through these key strengths:
Periodic Labs rides the AI-for-science wave, timing perfectly with post-AGI hype shifting to specialized, grounded applications amid compute abundance and robotics advances.[2][3] Physical sciences offer rich, verifiable data for AI training, unlike ambiguous fields, fueling progress in Moore’s Law extension, nuclear fusion, space travel, and semiconductors—critical as chip wars and energy transitions intensify.[2][3] Market forces like Nvidia's hardware dominance (as an investor) and Bezos/Schmidt's deep tech bets amplify this, while talent exodus from Big Tech labs creates a moat.[1][2] By generating exclusive experiment data, Periodic Labs influences the ecosystem as a foundational data flywheel, enabling broader AI R&D automation and reducing global innovation timelines from years to days.[1][3]
Periodic Labs is primed to dominate AI-native materials discovery, scaling autonomous labs to run thousands of daily experiments and deploy agent tools across pharma, energy, and semis.[1][2] Next steps include team growth, lab expansion via $300M war chest, and academic grants to fuel bold research—potentially unlocking fusion reactors or next-gen chips.[3] Trends like agentic AI, robotic dexterity, and sovereign AI labs will propel them, evolving their influence from niche partnerships to ecosystem enabler, redefining science as an always-on, AI-human hybrid. This positions them as the bridge from hype to hardware reality, accelerating breakthroughs that power the next tech era.[1][2][3]