A breakthrough deterministic physics kernel delivers molecular, materials, and reaction screening across three scientific domains from one unified engine.
OLBIA, SARDINIA, ITALY, March 20, 2026 /EINPresswire.com/ — FluxMateria today announced the public launch of FluxMateria.com and research-preview availability of its computational screening platform for scientific and industrial R&D teams.
FluxMateria is a fundamentally new approach to computational screening. It is neither built on density functional theory (DFT) nor on machine learning or AI. Instead, the platform is powered by a novel, deterministic physics kernel derived from a first-principles geometric framework with no training data. This single engine evaluates molecular properties, materials properties, reaction pathways, and spectroscopy, delivering the interpretability and reproducibility of physics-based methods at speeds previously achievable only by opaque AI models.
The public launch includes published benchmarks, no-signup interactive demos, guided walkthroughs, and pilot-access pathways for organizations to evaluate FluxMateria against their own workflows and datasets.
A New Category of Computational Tool
Computational chemistry and materials science have long faced a core tradeoff: physics-based methods like DFT are rigorous and interpretable but too slow for large-scale screening; AI/ML models are fast but lack explainability, struggle with out-of-distribution chemistry, and offer no way to know when they are guessing.
FluxMateria’s physics kernel breaks this tradeoff. Derived from first principles rather than fitted to data, it generalizes to novel chemistry immediately. Being fully deterministic, every result is traceable and reproducible. Operating at speeds up to 3.6 million times faster than conventional DFT, it makes exhaustive screening practical for the first time.
Headline Benchmarks
Life Sciences. Full ADMET panel (solubility, permeability, CYP inhibition, hERG, hepatotoxicity) at approximately 350 molecules per second (single-threaded, no GPU), validated across more than 175,000 compounds. Mechanism-of-action prediction 91% accurate across more than 10,000 targets. Binding affinity Pearson r = 0.77 on CASF-2016 (zero calibration).
Materials Science. Band-gap prediction MAE under 0.7 eV across more than 1,000 materials (metals, semiconductors, perovskites, TMDs). Bond-length prediction under 0.1% mean error across more than 450 bonds and 60+ elements.
Chemistry. 100% mechanism classification accuracy across 336 experimental cases; activation barrier MAE 7.4 kJ/mol. Solvation, synthesis planning (29 reaction types), and spectroscopy (IR, NMR, UV-Vis); all from the same kernel.
Every prediction includes a built-in confidence indicator (high, medium, or low) so teams know exactly where experimental follow-up is most valuable.
Full benchmark methodology and test conditions are published on the FluxMateria platform.
Platform Architecture
FluxMateria offers 11 computational modules through an API-first architecture (150+ endpoints). Enterprise features include role-based access, append-only audit logs with full provenance, organization-level data isolation, and usage-based billing.
“Scientific teams should not have to choose between speed, interpretability, and reproducibility,” said Roberto Campus, serial entrepreneur and founder of FluxMateria. “We built a new physics kernel. Not a faster DFT, not another AI black box. When the cost of asking a safety or screening question drops to near zero, the entire discovery pipeline changes: ADMET profiling moves to the beginning, materials screening becomes exhaustive instead of selective. That’s the real shift.”
FluxMateria is currently in research preview via live demos, pilot collaborations, and direct engagement with research teams. Initial focus areas include drug-discovery support, materials screening, reaction analysis, and enterprise platform evaluation.
Organizations can explore the public site, run no-signup demos, or request tailored pilot access at https://fluxmateria.com.
Media Contact
Roberto Campus
FluxMateria
contact@fluxmateria.com
About FluxMateria
FluxMateria is a computational discovery platform for life sciences, materials science, and chemistry. Built on a breakthrough physics kernel: not DFT, not AI. It evaluates molecular, materials, and reaction properties from a single deterministic engine with no training data, preserving full interpretability, traceability, and reproducibility with confidence indicators on every prediction.
Forward-Looking Statements
This press release contains forward-looking statements regarding product capabilities, planned workflows, pilot programs, and platform development. Actual results, availability, and product features may differ as the platform evolves during the research-preview period.
Roberto Campus
Fluxmateria Labs Srls
roberto@fluxmateria.com
Visit us on social media:
LinkedIn
Legal Disclaimer:
EIN Presswire provides this news content “as is” without warranty of any kind. We do not accept any responsibility or liability
for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this
article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
