Corollary

Research

  • Papers

Library

  • Catalog

Account

  • Jobs
  • Billing
SettingsResearch demo · not for clinical or commercial use
Corollary
  1. Catalog
  2. Chemistry & ADMET
  3. MegaMolBART

Chemistry & ADMET · NVIDIA / BioNeMo

MegaMolBART

Autoregressive SMILES generation and forward reaction prediction.

Commercial use permittedThe Apache-2.0 licence allows use in commercial research and products.Apache-2.0
Availability
Not runnable yet

No backend is configured for this model.

Typical latency
5.0s
Credits per run
~42
Compute tier
A · Fast

1–30s

Not runnable yet

This model isn't available to run right now.

Run this model

MegaMolBART

Canonical SMILES for the small molecule.

up to 105 credits · ~5s typical

About this model

Trained on ZINC and ChEMBL, MegaMolBART handles both unconditional generation and reaction product prediction, which makes it useful in synthesis planning as well as design.

Standardized I/O contract

Every model in the Hub speaks the same contract, which is what lets the Router and the agent call any of them without special-casing.

Inputs

  • smilessmiles

    Ligand SMILES

Outputs

  • moleculestable

    Sampled molecules.

Specification

Hardware
1× A100 16GB
GPU memory
16 GB
Version
1.0
Licence
Apache-2.0
Backend
not configured
MCP server
mcp-molecule-server

Tasks

molecular generationreaction prediction
generationreactionsnvidia