Protenix vs Boltz

Protenix: Trainable PyTorch reproduction of AlphaFold 3 by ByteDance. Boltz: Open biomolecular interaction models (Boltz-1, Boltz-2) for complex structure and binding-affinity prediction.

Facts below come from each project's own website, repository and paper. We don't publish benchmark claims here; see the papers for evaluations. Shared tags: structure.

FieldProtenixBoltz
TaskStructure prediction (trainable AlphaFold 3)Structure + binding-affinity prediction
ModalitiesProtein, ligand, nucleic acidBiomolecular complexes
MSA requiredYes (protein + RNA MSA, templates)Optional (MSA server flag)
Open weightsv1.0 open (368M); v2 (464M) proprietaryYes (MIT)
GPU need—Recommended; CPU supported but much slower
LicenseApache-2.0MIT
PricingOpen sourceOpen source
Released2024-12-122024-11-17
PaperLinkLink
GitHubGitHubGitHub
GitHub stars2,0804,254
InterfacePackage, WebPackage

License and commercial use

FieldProtenixBoltz
Code licenseApache-2.0MIT
Weights license—MIT
Commercial usePermissive (commercial OK)Permissive (commercial OK)
AnswerYes. The recorded license is permissive, so closed-source commercial use is allowed if you keep the notices. DetailsYes. The recorded license is permissive, so closed-source commercial use is allowed if you keep the notices. Details

Add more Protein structure & design tools to this comparison →

More Protein structure & design comparisons

← All Protein structure & design tools