# brannondorsey/PassGAN

A Deep Learning Approach for Password Guessing (https://arxiv.org/abs/1709.00440)

Repository: https://github.com/brannondorsey/PassGAN
Canonical: https://ross.abutalabs.com/products/passgan
Language: Python
License: MIT
License Family: permissive
Topics: password-cracking, machine-learning, deep-learning, gan, password, password-strength
Last push: 2023-02-24T07:01:06+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3179, "days_push": 1286, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2012, forks 399 (observed 2026-08-28T04:06:05.194110+00:00)

## What it is
A Python command-line implementation of the PassGAN paper that uses a Wasserstein GAN to generate password guesses, with a pretrained model trained on the RockYou dataset. It supports both sampling passwords from the pretrained checkpoint and training new models on custom password leak datasets.

## Use cases
- generate password guesses with a GAN for cracking hashed passwords
- audit password strength against ML-generated guesses
- reproduce the PassGAN research paper results
- train a password generation model on my own leaked password dataset
- sample millions of realistic passwords from a pretrained model

## When to choose
- you want to run GAN-based password guessing with a pretrained RockYou model
- you need a CLI to train or sample from a password-generating GAN
- you are reproducing or extending the PassGAN paper

## When to avoid
- you need a fast, production-grade password cracker - traditional rule/Markov tools like hashcat are faster
- you lack a CUDA GPU, since training requires one
- you want actively maintained TensorFlow code - it depends on old CUDA 8-era tooling

## Facets
- artifact type: cli-tool
- maturity: maintenance
- function: machine-learning, deep-learning, cli, security
- domain: security, machine-learning, deep-learning
- platform: python, cli
- tags: gan, password-cracking, password-guessing, wgan, research-reproducibility, pretrained-model, linux, gpu

## Member repositories
- brannondorsey/PassGAN (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:05.194110+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 2026-08-30T03:01:01.802172+00:00, confidence not recorded.
  - readme: https://github.com/brannondorsey/PassGAN (fetched 2026-08-28T04:06:05.194110+00:00, sha bafef37a388e)
- Data as of 2026-08-30T08:39:29.467469+00:00.