FLUXMATERIA — MATERIALS
Screen materials at the speed of ideas.
Input compositions. Get electronic, thermal, and structural property predictions. Shortlist candidates without waiting for DFT queues.
FLUXMATERIA — MATERIALS
Input compositions. Get electronic, thermal, and structural property predictions. Shortlist candidates without waiting for DFT queues.
Electronic structure and band gap calculations from composition
Thermal conductivity and stability estimates
Lattice parameters and structural stability
Direct composition-to-property predictions
Screen entire material libraries efficiently
Apply property constraints to shortlist candidates
Curie temperature predictions from composition — 7.2% MAPE on 79 materials
A dedicated battery-native decision layer for cathodes and build-ready battery screening. FluxMateria now combines bulk properties, transport, degradation, electrolyte and coating fit, uncertainty, and prototype handoff in one local workflow.
Topology-aware transport, bottleneck screening, anisotropy, defect sensitivity, and rate-capability signals.
Fade-mode breakdown, impedance growth, cracking pressure, voltage-window stress, and cycle-life heuristics.
Interphase risk, oxygen-loss pressure, dissolution risk, and coating recommendations tied to the chemistry.
Cost competitiveness and manufacturing readiness folded directly into the battery decision layer.
Support-aware confidence, out-of-domain warnings, and experiment recommendations for the next iteration.
Build-facing handoff with recommended configuration, validation plan, and public/internal disclosure split.
See how the module changed the answer from bulk winner to interface winner to build winner in one local run.
Read case studyDownload the public writeup of the battery workflow, results, and literature context.
Open white paperReview holdout accuracy and the multi-scenario stress test for the battery layer.
Open benchmark| Approach | Time per material | Practical for screening? |
|---|---|---|
| Full DFT | Hours to days | No |
| ML surrogates | Seconds | Yes, but limited extrapolation |
| FluxMateria | Seconds | Yes, physics-based |
This table is a positioning summary. Measured throughput and accuracy are reported on the Benchmarks page.
Enter composition or upload material library
Choose property predictions (electronic, thermal, structural)
Fast batch computation
Apply property thresholds
Review ranked candidates meeting criteria
Decision packet for DFT follow-up
Validated against DFT reference calculations and experimental measurements across material families.
| Property | Metric | FluxMateria | DFT Reference | Status |
|---|---|---|---|---|
| Band Gap | MAE | 0.9% MAE | — | ✓ DFT-competitive |
| Metal vs Insulator | Accuracy | 100% | ~95% | ✓ Pass |
| Formation Energy | MAE | 0.08 eV/atom | — | ✓ Screening-grade |
| Lattice Parameters | MAPE | 2.1% | 1-2% | ✓ Competitive |
| Thermal Conductivity | Correlation | r² = 0.87 | r² ~ 0.9 | ✓ Screening-grade |
| Curie Temperature | MAPE | 7.2% | 15–30% | ✓ 79 materials, 17 families |
All predictions are physics-based. No ML black boxes. 100% interpretable.
No account required. Enter a composition and see property predictions with full interpretability.
Run Demo →Upload a material library and screen the full composition space.
Request Pilot Access →Patent Pending