# albertan017/LLM4Decompile

Reverse Engineering: Decompiling Binary Code with Large Language Models

Repository: https://github.com/albertan017/LLM4Decompile
Canonical: https://ross.abutalabs.com/products/llm4decompile
Homepage: https://aclanthology.org/2024.emnlp-main.203
Language: Python
License: MIT
License Family: permissive
Topics: decompile, reverse-engineering, large-language-models, binary
Last push: 2026-02-12T03:02:03+00:00

## Health v2 (maintenance only)
Score: 55/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 67, release rhythm 35, longevity 65
- inputs: {"age_days": 917, "days_push": 202, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6986, forks 548 (observed 2026-08-28T04:09:51.910694+00:00)

## What it is
LLM4Decompile is an open-source series of large language models (1.3B to 33B) trained to decompile binary code back into readable, executable high-level source code, outperforming traditional tools like Ghidra. The repository includes model weights, training scripts, and the HumanEval-Decompile and decompile-bench datasets for training and evaluation.

## Use cases
- decompile binary executables back into C source code
- recover readable source from stripped or optimized binaries
- improve on Ghidra's unreadable decompiler output with LLMs
- train a custom decompilation model on binary-source function pairs
- benchmark LLM decompilation with HumanEval-Decompile
- analyze malware or legacy binaries without source code

## When to choose
- you need ML-based decompilation with more readable and re-executable output than Ghidra
- you want open-source decompiler models you can fine-tune or self-host
- you are researching LLM applications in reverse engineering
- you need large binary-source paired datasets for training

## When to avoid
- you need a production-grade, guaranteed-correct decompiler for security audits
- you lack GPU resources to run multi-billion-parameter models
- you need decompilation for languages or architectures outside the training data
- you want a lightweight traditional decompiler without ML overhead

## Facets
- artifact type: library
- maturity: active
- function: reverse-engineering, machine-learning, llm-inference, llm-training, parser
- domain: reverse-engineering, large-language-models, security, machine-learning
- platform: python
- tags: decompilation, binary-analysis, ghidra, disassembly, research-models, huggingface, linux, gpu

## Member repositories
- albertan017/LLM4Decompile (main) score 55

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:51.910694+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 2026-08-29T17:41:07.945778+00:00, confidence not recorded.
  - readme: https://github.com/albertan017/LLM4Decompile (fetched 2026-08-28T04:09:51.910694+00:00, sha a5719585b86f)
  - homepage: https://aclanthology.org/2024.emnlp-main.203 (fetched 2026-08-29T08:37:03.480651+00:00, sha d67a70788f7f)
  - site_page: https://aclanthology.org/faq/news (fetched 2026-08-29T08:37:03.483467+00:00, sha 349322ec8ae9)
  - site_page: https://aclanthology.org/faq/related-work (fetched 2026-08-29T08:37:03.485122+00:00, sha a969c9a3324e)
  - site_page: https://aclanthology.org/faq/copyright (fetched 2026-08-29T08:37:03.486598+00:00, sha 6f52ca1d08e3)
  - site_page: https://aclanthology.org/faq/volunteer (fetched 2026-08-29T08:37:03.488038+00:00, sha db98d9a3c971)
  - site_page: https://aclanthology.org/faq/feedback (fetched 2026-08-29T08:37:03.489553+00:00, sha bf6cb8ae2675)
  - site_page: https://aclanthology.org/faq/bib (fetched 2026-08-29T08:37:03.491071+00:00, sha 997a5ef919bd)
  - site_page: https://aclanthology.org/faq/linking (fetched 2026-08-29T08:37:03.492643+00:00, sha 40d9799179cc)
  - site_page: https://aclanthology.org/faq/api (fetched 2026-08-29T08:37:03.494605+00:00, sha b1a063af84d5)
- Data as of 2026-08-30T08:39:29.467469+00:00.