Every atomchanges the metal.
A hands-on laboratory for materials that don’t exist yet. Mix elements, stretch lattices and send an agent hunting for candidates — and see which numbers are physics, which are rules of thumb, and which are guesses.
Look closer. Think wider.
Three habits for exploring a crystal without rushing to a simple answer.
Notice the atoms
Every element brings its own size, mass and electrons. Size mismatch is where strength — and trouble — begins.
Follow the bonds
Change one ingredient and watch the whole crystal respond, not only the site you touched.
Keep the uncertainty
Use the agent to compare trade-offs — never to declare a material discovered.
Lattice exploration
Choose a study, move the dial, then inspect what changes. Motion is visual only; every result stays reproducible.
Select any atom in the specimen to open its tiny field note.
Let the agentweigh the trade-offs.
Start the run to compare all eight settings.
From prompt to blueprint
Define properties
State the density, melting point or structure you are after.
Generate structures
The agent proposes compositions and a likely crystal arrangement.
Run estimates
Rule-of-mixtures and phase heuristics score each candidate.
Optimize
Keep the best, discard the rest, and show the trade-offs.
Publish a blueprint
Save the composition with its assumptions attached.
Alloid is a computational platform, not a source of experimentally validated materials. Promising candidates need physical synthesis and independent testing to confirm any property shown here.
Bring a property.Leave with a candidate.
Describe the material you need. The discovery agent searches hundreds of compositions and prints its assumptions next to every result.
Deploy the discovery agentMaterial library
Real materials, real mechanisms. Documented findings first, model guesses second.
Educational summaries, not lab records. Documented findings and illustrative model outputs are labeled separately.
Describe a material.Question the answer.
What should it do?
What matters most?
Which elements may it use?
Good discovery leaves room for doubt.
Each candidate is scored by simple, published rules of thumb. Your report separates what those rules estimate from what only a furnace can tell you.
- Whether brittle intermetallic compounds form — that needs CALPHAD or DFT calculations.
- Real strength, toughness or creep life, which depend on processing and microstructure.
- Oxidation and corrosion behaviour beyond the presence of oxide-forming elements.
- Cost, supply risk and whether the alloy can be cast or machined at all.
Alloy forge
Elements. Ratios. Lattice. A learner-defined alloy, rule-of-mixtures estimates and the classic high-entropy alloy screening criteria — shown with their limits.
Density and melting point here are rule-of-mixtures estimates; real alloys can melt hundreds of kelvin lower and form phases these rules cannot see. The VEC thresholds follow Guo et al. (2011); the δ ≤ 6.6% screen follows Yang and Zhang (2012).
Lab notebook
Blueprints you save in the forge, ranked by a simple specific-melting index (melting estimate ÷ density). Stored only in this browser.
Research in.Rewards out.
Engineer materials
Users generate experimental alloys, crystals and composites.
Spend $ALLOID
Unlock advanced simulations, compute and autonomous agents.
Generate discoveries
Runs produce candidate structures with predicted properties.
Reward innovation
Proposed platform fees fund buybacks and rewards for reproducible results.
Token mechanics on this page are proposals, not commitments. Nothing here is financial advice, and computational results are not evidence that a material works.