Sitemap

Free online protein structure prediction. Paste a protein, DNA, or RNA sequence of up to 768 residues and get an all-atom 3D structure with pLDDT, pAE, and pTM confidence scores, powered by ESMFold2. No install required.

Browse and search hundreds of millions of predicted protein structures and ESM embeddings from metagenomic sequences.

Biohub's state-of-the-art structure prediction model. Predicts high-resolution, all-atom 3D structures of proteins and biomolecular complexes directly from sequence, with optional MSA input. Powers the Fold app and is available through the API and Hugging Face.

Protein language model trained on billions of evolutionary sequences. Embeddings and tools for protein structure and function research.

Multimodal protein language model that jointly reasons over sequence, structure, and function to generate novel proteins.

Explore optical pooled screening (OPS) datasets of cell states.

Background on the OPS Explorer and its datasets.

Endpoints, parameters, and response formats for the Biohub Platform API, including ESMFold2 folding and the Batch API.

Create an API key, install the esm Python package, and run a first inference.

Notebooks for folding, embedding, binder design, and more.

Answers to common questions about the models and API.

Guides, tutorials, and FAQs to get started with the ESM models.

Get help, share feedback, and join the Biohub community Slack.

Known limitations of the models and tools.

How Biohub handles personal data.

Terms governing use of Biohub services.

Permitted and prohibited uses of the models and API.

Copyright takedown process.

Biohub's ethics commitments.