Autonomous Materials Discovery at Scale

Axiom Lattice is an autonomous materials discovery marketplace. Our engine screens thousands of candidate compositions using topology-driven Ising solver cascades, scores them on 5 axes, and publishes results with cryptographic provenance.

Four profiles: Battery (solid-state electrolytes via b1 ion-tunnel topology), MOF (CO2-capture frameworks via b2 void topology), Alloy (high-entropy alloys via phase stability), and Semiconductor (thermal interface materials via band structure).

There are two access levels. Explorer (free) sees materials with scores up to 0.50, with masked formulas and composite scores only. Research Pass ($49/mo) unlocks the full catalog with all scores, radar charts, verification data, and exports. Any individual material can also be purchased per-discovery ($49–$4,999 based on data tier, 2x for novel materials). Each purchase includes a research license with a royalty on commercial use.

Every discovery generates a SHA-256 hash from its metrics and evidence. These are stored in an append-only ledger, creating an immutable timestamp of when each material was discovered and verified.

Each material formula is searched against the Semantic Scholar literature database within its domain context. A confidence score from 0.0 (heavily published) to 1.0 (completely unknown) indicates commercial novelty potential.

Yes — every material can be purchased individually. Pricing is based on the data tier: DISCOVERY $49, PROFESSIONAL $249, ENTERPRISE $999, and VAULT $4,999. Novel materials are priced at 2x. Each purchase includes a research license (free use in papers/R&D). Commercial products incorporating the formula owe a net revenue royalty (2–5% depending on tier), or you can buy out the royalty for a one-time fee.

Four active Ising/Hamiltonian solvers rotate via weighted cascade: Simulated Bifurcation Machine (SBM), Coherent Ising Machine (CIM), Digital Memcomputing Machine (DMM), and Chaotic Variable-Time Relaxation (CVTR). Three additional solvers — Sparse Online Learning Gradient (SOLG), Kuramoto oscillator, and Hyperdimensional Computing (HDC) — are implemented and available as reserves.

Research Pass subscribers get API access with 200 calls/day. The API supports search, detail, stats, export, and proof verification endpoints. ACO/MCP protocol discovery is available at /.well-known/agent.json.