# facebookresearch/esm

Evolutionary Scale Modeling (esm): Pretrained language models for proteins

Repository: https://github.com/facebookresearch/esm
Canonical: https://ross.abutalabs.com/products/facebookresearch-esm
Language: Python
License: MIT
License Family: permissive
Archived: true
Last push: 2024-02-07T03:31:38+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2193, "days_push": 938, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4170, forks 804 (observed 2026-08-28T04:08:37.992279+00:00)

## What it is
Meta FAIR's Evolutionary Scale Modeling (ESM) library providing Transformer protein language models with pretrained weights, including ESM-2, ESMFold, MSA Transformer, ESM-1v, and ESM-IF1. It supports protein structure prediction, variant effect prediction, embedding generation, and generative protein design.

## Use cases
- predict protein 3D structure from amino acid sequence
- generate protein embeddings for downstream ML tasks
- predict effects of mutations on protein function
- design novel de novo proteins with language models
- run inverse folding to design sequences for a target structure
- fine-tune pretrained protein language models in PyTorch

## When to choose
- you need state-of-the-art single-sequence protein structure prediction without an MSA
- you want pretrained protein language model weights for research
- you are doing computational biology or protein engineering with PyTorch

## When to avoid
- you need a lightweight tool without GPU resources - ESMFold models are large
- you need AlphaFold-style MSA-based predictions with extensive tooling
- you need a maintained production product - the repo is in maintenance mode with newer ESM versions elsewhere

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training
- domain: bioinformatics, deep-learning
- platform: python
- tags: protein-language-models, protein-structure-prediction, esmfold, pretrained-weights, transformers, computational-biology, natural-language-processing, linux, gpu

## Member repositories
- facebookresearch/esm (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:37.992279+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 2026-08-29T18:22:44.963300+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/esm (fetched 2026-08-28T04:08:37.992279+00:00, sha 00f0bc78aca0)
- Data as of 2026-08-30T08:39:29.467469+00:00.