NIM for Boltz2

Type: Microservice Tags: NVIDIA, NIM, BioNeMo, Boltz-2, protein structure, biomolecular complexes, binding affinity, ligands, drug discovery Related: NVIDIA-BioNeMo, NVIDIA-NIM, NIM-for-MSA-Search, NIM-for-OpenFold3, NIM-for-OpenFold2, NIM-for-AlphaFold2, NIM-for-AlphaFold2-Multimer, NIM-for-Evo-2, NIM-for-DiffDock, NIM-for-ProteinMPNN, cuEquivariance, TensorRT, NVIDIA-Clara, NVIDIA-AI-Enterprise, NGC Sources: https://docs.nvidia.com/nim/bionemo/boltz2/latest/overview.html; https://docs.nvidia.com/nim/bionemo/boltz2/latest/index.html Last Updated: 2026-04-29

Summary

NIM for Boltz2 is NVIDIA’s NIM microservice for Boltz-2 biomolecular structure prediction and binding-affinity prediction. It targets proteins, RNA, DNA, ligands, protein-nucleic-acid complexes, modified residues, ligand binding, and constraint-guided predictions for drug-discovery and structural-biology workflows.

Detail

The current docs describe Boltz-2 NIM as a biomolecular structure and binding-affinity prediction service for combinations of proteins, RNA, DNA, and other molecules. It supports both single-molecule predictions and complex multi-molecular assemblies, including protein structures, nucleic acid structures, protein-nucleic acid complexes, ligand binding and affinity scores, modified residues, and predictions conditioned on pockets or contacts.

Boltz2 belongs near NIM-for-OpenFold3 because both model biomolecular complexes. Boltz2 is especially important for ranking and assessing molecular interactions because current docs call out binding-affinity prediction as a first-class capability. It also belongs near cuEquivariance and TensorRT because NVIDIA positions BioNeMo structure models around optimized GPU kernels and inference engines.

Connections

Source Excerpts

  • “binding affinity prediction”
  • “proteins, RNA, DNA”
  • “constraint-guided predictions”