Accelerating Scientific Research with Gemini in the Real-World
Samuel Schmidgall and a large Google team extend Co-Scientist from an in-silico hypothesis generator into an execution-grounded research partner, validating it against real wet-lab and computational outcomes across materials science, biology and computer science.
Ask this paper
Closed loop with physical instruments: Co-Scientist interfaced with a semi-automated chemical vapor deposition reactor to design a safe precursor route for MXenes; execution produced a lamellar 2D material with key structural similarities to the Ti3C2Tx lattice, with atomic structure still unconfirmed.
Lab-in-the-loop recipe tailoring: Using Gemini 3 Deep Think, it tailored growth recipes to laboratory constraints in minutes, enabling single-attempt growth of monolayer MoS2, MoSe2 and WS2.
Quantitative biology prediction: It predicted emergent swarming phenotypes of engineered E. coli across IPTG gradients from sparse imaging data, matching unpublished wet-lab morphological measurements.
Self-discovered architecture: In computer science it autonomously found an inference-time scaling architecture that beat six frontier models on HealthBench Hard and Professional while reducing potential clinical harm under blinded physician evaluation.
Reviewed at scale: A double-blind study of end-to-end generated papers with 30 domain experts across 450 reviews found the reliability modules reduce hallucination and plagiarism while improving research safety.
Abstract
We present an extension and comprehensive real-world validation of Co-Scientist, a Gemini-based multi-agent system designed to accelerate end-to-end scientific research across hypothesis generation, experimentation, and manuscript generation. Moving beyond in silico hypothesis generation, this specialized configuration transitions Co-Scientist into an execution-grounded research partner advancing closed-loop scientific workflows across materials science, biology, and computer science. In materials science, Co-Scientist interfaced with a semi-automated chemical vapor deposition reactor to design a safe precursor route for MXenes; experimental execution produced a lamellar 2D material sharing key structural similarities with the Ti3C2Tx MXene lattice, although further experiments are needed to confirm the atomic structure. Leveraging Gemini 3 Deep Think for rapid, lab-in-the-loop execution, it also tailored growth recipes to laboratory constraints in minutes, enabling single-attempt growth of monolayer MoS2, MoSe2, and WS2 semiconductors. In biology, Co-Scientist predicted emergent swarming phenotypes of engineered E. coli across inducer (IPTG) gradients from sparse imaging data, quantitatively matching unpublished wet-lab morphological measurements. In computer science, Co-Scientist autonomously discovered an inference-time scaling architecture that outperformed six frontier models on HealthBench (Hard and Professional) while reducing potential clinical harm under blinded physician evaluation. Finally, a double-blind study of end-to-end generated papers with 30 domain experts across 450 reviews demonstrates that Co-Scientist's reliability modules reduce hallucination and plagiarism while improving research safety. Together, these results demonstrate progress toward closed-loop multi-agent scientific AI systems capable of accelerating real-world scientific discovery.