gpt_academic

GPT Academic provides a practical interaction interface for large language models such as GPT and GLM, with particular optimization for paper reading, polishing, and writing. It is modular and supports custom shortcut buttons and function plugins. The project supports project analysis and self-explanation for Python, C++, and other codebases, as well as PDF and LaTeX paper translation and summarization. It can query multiple LLM models in parallel and supports local models such as ChatGLM3. It also connects to models including Tongyi Qianwen, DeepseekCoder, iFlytek Spark, ERNIE Bot, LLaMa2, RWKV, Claude2, and MOSS.

The project includes built-in function plugins for one-click polishing, translation, finding paper grammar errors, and explaining code; one-click analysis of Python/C/C++/Java/Lua project trees or self-analysis; one-click interpretation of full LaTeX/PDF papers and summary generation; LaTeX full-text translation and polishing; batch comment generation; Markdown Chinese-English translation; PDF paper full-text translation; an Arxiv assistant; LaTeX paper proofreading; a Google Scholar integration assistant; and internet information aggregation with GPT. It also supports formula, image, and table display, dark theme, and multiple LLM model support.

Installation can be done by running directly on Windows, Linux, or MacOS, by using Docker, or through other deployment methods. The project also provides a Windows one-click run script for users unfamiliar with Python environments, and supports deployment via Sealos, WSL2, and subpath hosting. It is licensed under GPL-3.0.