# lucidrains/alphafold3-pytorch

Implementation of Alphafold 3 from Google Deepmind in Pytorch

Repository: https://github.com/lucidrains/alphafold3-pytorch
Canonical: https://ross.abutalabs.com/products/alphafold3-pytorch
Language: Python
License: MIT
License Family: permissive
Topics: artificial-intelligence, deep-learning, attention-mechanisms, denoising-diffusion, transformers, protein-structure-prediction
Last push: 2026-08-10T02:16:28+00:00

## Health v2 (maintenance only)
Score: 68/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 96, release rhythm 36, longevity 60
- inputs: {"age_days": 847, "days_push": 24, "days_rel": 349, "gap_med": 0, "n_releases_24m": 114}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1691, forks 229 (observed 2026-08-28T04:05:22.593268+00:00)

## What it is
A PyTorch implementation of Google DeepMind's Alphafold 3 model for protein and molecular structure prediction. It is a research-oriented reimplementation of the paper's architecture, including training pipeline components like PDB dataset preparation and sampling.

## Use cases
- predict protein structures from sequences in pytorch
- reproduce alphafold 3 training from scratch
- research alphafold 3 architecture components
- train a structure prediction model on PDB data
- experiment with diffusion-based molecular structure prediction

## When to choose
- you want a pure PyTorch, hackable implementation of Alphafold 3 for research
- you need to study or extend the AF3 architecture in code
- you want to train or fine-tune structure prediction models on your own data

## When to avoid
- you need a production-ready, officially supported inference tool with pretrained weights
- you need optimized performance - consider alternatives like MegaFold with Triton kernels
- you need turnkey protein folding predictions without ML expertise

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-training
- domain: deep-learning, bioinformatics, artificial-intelligence
- platform: python
- tags: alphafold3, protein-structure-prediction, pytorch, diffusion, transformers, research-implementation, gpu

## Member repositories
- lucidrains/alphafold3-pytorch (main) score 68

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:22.593268+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 2026-08-30T03:38:16.379196+00:00, confidence not recorded.
  - readme: https://github.com/lucidrains/alphafold3-pytorch (fetched 2026-08-28T04:05:22.593268+00:00, sha a444c26a622f)
- Data as of 2026-08-30T08:39:29.467469+00:00.