The computational chemistry engine AI agents call to do real discovery.
Pre-computed molecular intelligence, GPU simulation, and quantum chemistry across 122 million compounds. Accessible to any MCP-compatible assistant. Swap the model freely. The engine stays.
Bringing one drug to approval costs about $2.6 billion, and roughly one in twenty survives. We move the decision in silico, before the spend.
pre-enriched
each
response
tools
servers
jurisdictions
Models are interchangeable. The engine they run on is not.
Remove the model and you swap one assistant for another. Remove NovoMCP and you lose 122M molecules, 69 validated tools, compliance, the audit trail, and cross-run memory. When the model is free to replace and the engine underneath is not, the engine is where the value lives.
One intelligence layer. One simulation engine.
Novo holds the pre-computed chemistry your AI reasons over. Novo Compute runs the physics when a question needs it. Together they let an assistant think, test, and learn about molecules without waiting.
The intelligence layer
Pre-computed ADMET, compliance, target evidence, and discovery orchestration. Answered in milliseconds, not job queues.
- →122M compounds, 100+ fields each
- →NovoExpert ADMET and clinical-outcome models
- →FAVES compliance at the point of decision
- →12-stage autonomous discovery funnel
The simulation engine
GPU and quantum chemistry for the questions that need real physics. Docking, dynamics, and quantum methods on demand.
- →AutoDock-GPU docking with strain correction
- →GROMACS molecular dynamics
- →xTB, CREST, and neural-network potentials
- →OpenFold3 structure prediction
The scientific method, run end to end.
One instruction starts a twelve-stage discovery funnel. Targets are found and validated, the literature read, actives pulled, candidates profiled, optimized, docked, and gated on a calibrated clinical-clearance estimate — then carried through molecular dynamics, optional FEP, and patient stratification.
Every stage writes to an immutable audit trail. Cross-run memory carries what prior runs learned into the next.
How AgentMode worksDepth in the fields chemistry decides.
Not a claim to every industry. Four domains where the tools already ship and the accuracy is benchmarked.
Drug discovery narrows millions to one candidate through a twelve-stage funnel. Materials work stays a flexible toolkit — four workflows, no forced pipeline. One engine underneath both: xTB, CREST, NNPs, FEP, reaction thermodynamics, transition states. Same physics, two shapes.
Before you govern it, you can see it.
Every decision the AI makes is written to an immutable audit trail. You can reconstruct any funnel: what was tried, what was rejected, why, and who approved it.
Observability is not bolted on after the fact. It is built into the engine, and it is the first question every enterprise buyer asks.
Immutable audit trail
Per-stage logging across the full discovery funnel. Reproducible, reconstructable, exportable.
FAVES V4 compliance
1,585 SMARTS across 8 jurisdictions. Runs inline at the point of decision, not as a terminal gate.
Cross-run memory
Funnel context persists across sessions. The engine remembers what prior runs learned.
Agent governance
RFC 8693 token exchange, per-agent key scoping, and agent-level audit on Enterprise.
The work, published.
Outcome-level methods and benchmarks, in the open.
From the newsroom.
The next discovery runs on an engine you can trust.
NovoMCP is open to a small group of PIs, postdocs, and research engineers. Tell us what you are working on.