---
title: N64LLMDecompile
emoji: 🎮
colorFrom: indigo
colorTo: purple
sdk: gradio
app_file: app.py
pinned: false
license: apache-2.0
tags:
  - build-small-hackathon
  - backyard-ai
  - thousand-token-wood
  - best-demo
  - best-use-of-modal
  - best-use-of-codex
  - track:backyard
  - track:wood
  - sponsor:openai
  - sponsor:nvidia
  - sponsor:modal
  - achievement:offgrid
  - achievement:welltuned
  - achievement:fieldnotes
---

# N64LLMDecompile 🎮

Decompile Nintendo 64 MIPS functions back into C with a fine-tuned LLM, with
compiler-verified scoring against the original target objects.

Built for the Hugging Face **Build Small** hackathon.

**Links:**
- 💻 Source code: [github.com/MatthewLReingold/N64LLMDecompile](https://github.com/MatthewLReingold/N64LLMDecompile)
- 🎥 Demo video: **[ YouTube link ](https://youtu.be/GVNjPEQKtAY)**
- 📝 Writeup: **[ blog link ](https://huggingface.co/blog/MatthewReingold/n64-decomp-dev-blog)**
- 📣 Social post: **[ LinkedIn Post](https://www.linkedin.com/posts/matthew-l-reingold_i-spent-the-huggingface-build-small-hackathon-share-7472408495703900160-h__B/?utm_source=share&utm_medium=member_desktop&rcm=ACoAAB3-BzsBKnd6prFEjrTrnAqYnSdX6ZF6egc)**

## What it does

Given the Ghidra pseudo-C of an N64 MIPS function, the model rewrites it into C
source. Optionally, the Space compiles that C with the original compiler and flags
and diffs the result against the target object using [asm-differ] — giving a
byte-level match score, where `0` means a byte-perfect match (the metric the decomp
community uses to confirm a function is "matched").

**Two ways to use it:**
- **Paste Ghidra pseudo-C** → get C back.
- **Also upload the target object** and pick the compiler + flags → get a match score.

## The model

[`MatthewReingold/llm4decompile-9b-v2-n64-finetune`][model] — a fine-tune of
[LLM4Decompile-9B-v2][base] on verified N64 scratches from decomp.me.

## Results & writeup

_Full results, methodology, and what I learned are covered in the blog post:_

**[ writeup link coming soon ]**

## Acknowledgments

Thank you to the work of these teams who all gave the resources used during development:

- **[decomp.me]** — the scratch database and compiler-toolchain packaging the dataset is built from
- **[decompals]** — IDO static recompilation, and the MIPS GCC / binutils builds
- **Paper Mario team ([pmret])** — the GCC 2.8.1 compiler
- **[LLM4Decompile][base] (LLM4Binary)** — the base model this fine-tune is built on
- **[asm-differ]** (simonlindholm) — the assembly diffing used for scoring
- **[Ghidra]** (NSA) — pseudo-C generation in the data pipeline
- **NVIDIA** — the Nemotron models, used during experimentation
- **[Modal]** — cloud GPU infrastructure and compute credits
- **OpenAI** — Codex, used during development
- **Hugging Face** — hosting the hackathon and ZeroGPU access

## License

Apache-2.0. The base model and the decomp toolchains retain their own respective licenses.

[model]: https://huggingface.co/MatthewReingold/llm4decompile-9b-v2-n64-finetune
[base]: https://huggingface.co/LLM4Binary/llm4decompile-9b-v2
[asm-differ]: https://github.com/simonlindholm/asm-differ
[decomp.me]: https://decomp.me
[decompals]: https://github.com/decompals
[pmret]: https://github.com/pmret
[Ghidra]: https://github.com/NationalSecurityAgency/ghidra
[Modal]: https://modal.com