Apps · Reverse-engineering plugin for Ghidra (needs a local model server)

OGhidra

by Lawrence Livermore National Laboratory · See repo (LLNL) · 466 GitHub stars

Chat with Ghidra in plain English: an Ollama model runs multi-step analysis through Ghidra’s tools and reports what it found.

macOSWindowsLinux

Get OGhidra

At a glance

Best for
Fully offline binary analysis with Ollama, without setting up a separate MCP client.
Inference engine
  • uses Ollama
GPU support
None of its own: it talks to a llama.cpp or MLX LM server
Opens
  • GGUF
  • MLX
Features
  • Chat window
  • Tool calling
  • Agent workflows
Platforms
Python 3.12+ with Ghidra 12.0.3 and Java 21.

Sources:

Install

macOS

git clone https://github.com/llnl/OGhidra && cd OGhidra && pip install -r requirements.txt

Windows

git clone https://github.com/llnl/OGhidra && cd OGhidra && pip install -r requirements.txt

Linux

git clone https://github.com/llnl/OGhidra && cd OGhidra && pip install -r requirements.txt

Then build and install its Ghidra plugin, pull a model with Ollama (its README suggests gemma3:27b to start and gpt-oss:120b for best results) and set OLLAMA_MODEL.

5 models on this site run in OGhidra

Models tagged for reverse engineering that support tool calling, with a GGUF or MLX build to serve locally. Raw decompilers such as LLM4Decompile run on their own in llama.cpp or LM Studio instead.

Model Made by Type Size Opens it as Quality tier
GLM-4.7 Flash (30B-A3B) Zhipu AI LLM 31B (3.0B active) GGUF or MLX 88
Devstral 2 (123B) Mistral AI LLM 125B GGUF or MLX 86
Qwen3 Coder 30B-A3B Alibaba LLM 31B (3.3B active) GGUF or MLX 78
Qwen3 30B-A3B Alibaba LLM 31B (3.3B active) GGUF or MLX 74
gpt-oss 20B OpenAI LLM 21B (3.6B active) GGUF or MLX 70

Alternatives

Which app should I use?

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.

Was this page helpful?