# jwohlwend/boltz

Official repository for the Boltz biomolecular interaction models

Repository: https://github.com/jwohlwend/boltz
Canonical: https://ross.abutalabs.com/products/boltz
Language: Python
License: MIT
License Family: permissive
Topics: binding-prediction, drug-discovery, protein-structure
Last push: 2026-05-29T08:20:25+00:00

## Health v2 (maintenance only)
Score: 63/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 84, release rhythm 47, longevity 46
- inputs: {"age_days": 654, "days_push": 96, "days_rel": 359, "gap_med": 5.0, "n_releases_24m": 17}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4177, forks 884 (observed 2026-08-28T04:08:38.259331+00:00)

## What it is
Boltz is a family of open-source biomolecular interaction models (Boltz-1 and Boltz-2) that predict complex structures and binding affinities of proteins, ligands, nucleic acids, and other molecules. It approaches AlphaFold3 accuracy and enables fast in silico screening for early-stage drug discovery.

## Use cases
- predict protein-ligand complex structures
- estimate binding affinity of small molecules to protein targets
- screen compound libraries for binders in early drug discovery
- cofold protein complexes with nucleic acids and ligands
- find an open-source alternative to AlphaFold3
- run batched structure predictions from YAML inputs

## When to choose
- you need open-source (MIT) biomolecular structure and affinity prediction for academic or commercial use
- you want near-FEP accuracy binding affinity estimates at much lower compute cost
- you need GPU-accelerated batch inference of biomolecular complexes

## When to avoid
- you need physics-based free-energy perturbation precision for lead optimization
- you lack GPU hardware and need fast predictions, since CPU inference is significantly slower
- you need a polished GUI or web service rather than a Python/CLI workflow

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-inference
- domain: bioinformatics, artificial-intelligence, healthcare
- platform: python, cli
- tags: protein-structure-prediction, binding-affinity, drug-discovery, alphafold, molecular-modeling, cofolding, linux, gpu

## Member repositories
- jwohlwend/boltz (main) score 63

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:38.259331+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:22:41.219480+00:00, confidence not recorded.
  - readme: https://github.com/jwohlwend/boltz (fetched 2026-08-28T04:08:38.259331+00:00, sha 45eceea60f3b)
- Data as of 2026-08-30T08:39:29.467469+00:00.