Open Source Python Research Software
Sort By:
Python Research Software
View 85 business solutionsBrowse free open source Python Research Software and projects below. Use the toggles on the left to filter open source Python Research Software by OS, license, language, programming language, and project status.
-
$300 Free Credits to Build on Google CloudPut your $300 in credit toward real workloads, then keep building with free monthly usage for 20+ products. No commitment and no charge until you upgrade.
-
Go from Code to Production URL in SecondsSkip the Kubernetes configs. Cloud Run handles HTTPS, scaling, and infrastructure automatically. Two million requests free per month.
-
1
Megatron-LM
Ongoing research training transformer models at scale
Megatron-LM is a GPU-optimized deep learning framework from NVIDIA designed to train extremely large transformer-based language models efficiently at scale. The repository provides both a reference training implementation and Megatron Core, a composable library of high-performance building blocks for custom large-model pipelines. It supports advanced parallelism strategies including tensor, pipeline, data, expert, and context parallelism, enabling training across massive multi-GPU and multi-node clusters. The framework includes mixed-precision training options such as FP16, BF16, FP8, and FP4 to maximize performance and memory efficiency on modern hardware. Megatron-LM is widely used in research and industry for pretraining GPT-, BERT-, T5-, and multimodal-style models, with tooling for checkpoint conversion and interoperability with Hugging Face. Overall, it is a production-grade system for organizations pushing the limits of large-scale language model training. -
2
Local Deep Research
95% on SimpleQA (e.g. Qwen3.6-27B on a 3090)
Local Deep Research is an open-source AI-powered research assistant designed to perform deep, iterative investigations by combining large language models with multi-source search capabilities. It runs locally, giving users full control over their data, privacy, and infrastructure while supporting both local and cloud-based LLMs. The system breaks down complex queries into smaller steps, performs parallel searches across web and academic sources, and generates structured, citation-backed reports. It also supports personal document ingestion through vector search, enabling users to build a private, searchable knowledge base. The platform includes a web interface, Docker-based deployment, and flexible configuration options, making it accessible to both developers and researchers. Its architecture emphasizes transparency, customization, and reproducibility in AI-assisted research workflows. -
3
AutoResearchClaw
Autonomous research from idea to paper. Chat an Idea. Get a Paper 🦞
AutoResearchClaw is an open-source framework designed to automatically generate full academic research papers from a single idea or topic. Built in Python, it orchestrates a multi-stage research pipeline that gathers literature, formulates hypotheses, runs experiments, analyzes results, and writes the final paper. The system retrieves real academic references from sources such as arXiv and Semantic Scholar to ensure credible citations. It can automatically generate code for experiments, run them in a sandbox environment, and analyze the results with statistical methods. The platform also uses multi-agent debate and automated peer review processes to refine research findings and improve paper quality. By combining literature discovery, experimentation, and writing automation, AutoResearchClaw aims to turn research ideas into conference-ready papers with minimal human intervention. -
4
Catalyst
Accelerated deep learning R&D
Catalyst is a PyTorch framework for accelerated Deep Learning research and development. It allows you to write compact but full-featured Deep Learning pipelines with just a few lines of code. With Catalyst you get a full set of features including a training loop with metrics, model checkpointing and more, all without the boilerplate. Catalyst is focused on reproducibility, rapid experimentation, and codebase reuse so you can break the cycle of writing another regular train loop and make something totally new. Catalyst is compatible with Python 3.6+. PyTorch 1.1+, and has been tested on Ubuntu 16.04/18.04/20.04, macOS 10.15, Windows 10 and Windows Subsystem for Linux. It's part of the PyTorch Ecosystem, as well as the Catalyst Ecosystem which includes Alchemy (experiments logging & visualization) and Reaction (convenient deep learning models serving). -
Ship Agents FasterGemini Enterprise Agent Platform lets you rapidly build, scale, govern and optimize production-ready agents grounded in your organization's data. The platform enables developers to build custom or pre-built agents for virtually any use case. New customers get $300 in free credits.
-
5
DIG
A library for graph deep learning research
The key difference with current graph deep learning libraries, such as PyTorch Geometric (PyG) and Deep Graph Library (DGL), is that, while PyG and DGL support basic graph deep learning operations, DIG provides a unified testbed for higher level, research-oriented graph deep learning tasks, such as graph generation, self-supervised learning, explainability, 3D graphs, and graph out-of-distribution. If you are working or plan to work on research in graph deep learning, DIG enables you to develop your own methods within our extensible framework, and compare with current baseline methods using common datasets and evaluation metrics without extra efforts. It includes unified implementations of data interfaces, common algorithms, and evaluation metrics for several advanced tasks. Our goal is to enable researchers to easily implement and benchmark algorithms. -
6
Open Source Vizier
Python-based research interface for blackbox
Open Source (OSS) Vizier is a Python-based interface for blackbox optimization and research, based on Google’s original internal Vizier, one of the first hyperparameter tuning services designed to work at scale. Allows a user to setup an OSS Vizier Server, which can host black-box optimization algorithms to serve multiple clients simultaneously in a fault-tolerant manner to tune their objective functions. Defines abstractions and utilities for implementing new optimization algorithms for research and to be hosted in the service. A wide collection of objective functions and methods to benchmark and compare algorithms. Define a problem statement and study configuration. Setup a local server, setup a client to connect to the server, perform a typical tuning loop, and use other client APIs. -
7
nanoGPT
The simplest, fastest repository for training/finetuning models
NanoGPT is a minimalistic yet powerful reimplementation of GPT-style transformers created by Andrej Karpathy for educational and research use. It distills the GPT architecture into a few hundred lines of Python code, making it far easier to understand than large, production-scale implementations. The repo is organized with a training pipeline (dataset preprocessing, model definition, optimizer, training loop) and inference script so you can train a small GPT on text datasets like Shakespeare or custom corpora. It emphasizes readability and clarity: the training loop is cleanly written, and the code avoids heavy abstractions, letting students follow the architecture step by step. While simple, it can still train non-trivial models on modern GPUs and generate coherent text. The project has become widely used in tutorials, courses, and experiments for people learning how transformers work under the hood. -
8
Brain Tokyo Workshop
Experiments and code from Google Brain’s Tokyo research workshop
The Brain Tokyo Workshop repository hosts a collection of research materials and experimental code developed by the Google Brain team based in Tokyo. It showcases a variety of cutting-edge projects in artificial intelligence, particularly in the areas of neuroevolution, reinforcement learning, and model interpretability. Each project explores innovative approaches to learning, prediction, and creativity in neural networks, often through unconventional or biologically inspired methods. The repository includes implementations, experimental data, and supporting research papers that accompany published studies. Notable works such as Weight Agnostic Neural Networks and Neuroevolution of Self-Interpretable Agents highlight the team’s exploration of how AI can learn more efficiently and transparently. Overall, this repository serves as an open research hub for sharing ideas and advancing the understanding of intelligent systems. -
9
VulnX
Intelligent Bot, Shell can achieve automatic injection
vulnx, an intelligent Bot, Shell can achieve automatic injection, and help researchers detect security vulnerabilities in CMS systems. It can perform a quick CMS security detection, information collection (including sub-domain name, IP address, country information, organizational information and time zone, etc.), and vulnerability scanning. Vulnx is An Intelligent Bot Auto Shell Injector that detects vulnerabilities in multiple types of Cms, fast cms detection, information gathering, and vulnerability scanning of the target like subdomains, IP addresses, country, org, timezone, region, and more. Instead of injecting each and every shell manually as all the other tools do, VulnX analyses the target website checking the presence of a vulnerability if so the shell will be Injected by searching URLs with the dorks Tool. Detects CMS (wordpress, joomla, prestashop, drupal, opencart, magento, lokomedia). -
Custom VMs From 1 to 96 vCPUs With 99.95% UptimeLive migration and automatic failover keep workloads online through maintenance. One free e2-micro VM every month.
-
10
moebinv
C++ libraries for manipulations in non-Euclidean geometry
These are two C++ libraries for symbolic, numeric and graphical manipulations in non-Euclidean geometry. There is GUI which allows to interact with these libraries by mouse clicks. On a dipper level the first library Cycle implements basic operations on cycles (quadrics) through FSCc construction. The second library Figure operates on ensembles of cycles connected by Moebius-invariant relations, e.g. orthogonality. Both libraries are based on the Clifford algebra capacities of the GiNaC computer algebra system (http://ginac.de). Besides C++ libraries there is a Python wrapper, which can be used in interactive mode (https://codeocean.com/capsule/7952650/). Both libraries work in arbitrary dimensions and signatures of metric. Additionally, there are some 2D/3D-specific routines including a visualisation to PostScript files through Asymptote (http://asymptote.sourcefourge.net) software. The source is written in literate programming NoWeb. -
11
PS-Drone
Programming a Parrot AR.Drone 2.0 with Python - The Easy Way
The PS-Drone-API is a full featured SDK, written in and for Python, for Parrot's AR.Drone 2.0. It was designed to be easy to learn, but it offers the full set of the possibilities of the AR.Drone 2.0, including Sensor-Data (aka NavData), Configuration and full Video-support. The video function is not restricted to mere viewing, it is also possible to analyze video images data using OpenCV2. Obviously, the PS-Drone is perfect for teaching purposes; however, even the requirements for professional purposes can be satisfied. PS-Drone comes with a tutorial, explaining its most important commands and the drone's most important sensor values. The examples are easy to understand for people with little programming experience. A full list of commands and a description of all sensor data is available in a detailed documentation. It took several months to create PS-Drone, so it would be nice to get some donations for further development (e.g. Parrot's Bebop) and as a appreciation. -
12
Computer Glossary
is a rich dictionary containing multiple computer-related terms
Computer Glossary is a rich dictionary containing multiple computer-related terms, which is useful for both students and professionals. The program integrates an offline database and offers references and descriptions for each term. Computer Glossary can be installed in just a few simple steps and does not require special skills. The program comes with a simple interface and lets you easily search any term. The results will let you see the word's definition, references to other related terms in the dictionary, examples, and the source of the information. -
13
xrayutilities
a package with useful scripts for X-ray diffraction physicists
xrayutilities is a python package used to analyze x-ray diffraction data. It can support with performing diffraction experiments and used for common steps in the data analysis. It can read experimental data from several data formats (spec, edf, xrdml, ...); convert them to reciprocal space for arbitrary goniometer geometries and different detector systems (point, linear as well as area detectors); for further processing the data can be gridded (transformed to a regular grid). More detailed description as well as documentation can be found at webpage http://xrayutilities.sourceforge.io/. Downloads for windows can be found on http://pypi.python.org/pypi/xrayutilities Development is performed on github: https://github.com/dkriegner/xrayutilities -
14
ACORBA
Automated approach to measure root tip angles of Arabidopsis thaliana
Gravitropic response is studied in most of the laboratories working with Arabidopsis thaliana, for example, to detect new phenotypes in mutants. However, manual analysis of images and microscopy data are known to be subjected to human bias. This is particularly the case for manual measurements of root bending as the angle is set subjectively. In this context, it is essential to develop and use automated or semi-automated image analysis to produce faster, reproducible, and unbiased data. In this context, we developped ACORBA (Automatic Calculation Of Root Bending Angles), a fully automated software to measure root bending angle over time. -
15
SimpleElastix
Medical Image Registration Library
SimpleElastix is an extension of SimpleITK that comes with the elastix C++ image registration library. This makes state-of-the-art medical image registration really easy to do in languages like Python, Java, C# and R. -
16
Acoustic Research Tool (ART)
Acoustic Simulation Library for Frequency and Time Domain Simulations.
ART is a flexible simulation framework for wind instruments. It includes a growing library of modelling elements. So far bore discontinuities, branches, tone holes, cylindrical and conical tubes, Bessel horns and bent tubes are available for frequency domain modelling. In the time domain generic bidirectional propagation elements, scattering elements, fractional delays, convolution with reflection functions and general z-domain networks are available and can be described using MuParserX expressions. Cylindrical and conical ducts can also be defined based on their geometry. Available models and their parameters can be enumerated and combined to form simulators for complex acoustical structures. Parameters can be specified symbolically by expressions containing other parameter values or global variables. Dependencies between parameters are resolved at run time. However, MuParserX expressions are compiled at design time. Zero-delay loops are detected and reported. -
17
Libro
An interactive program for statistical analysis of texts
A cross-platform text analysis program written in Python and Free Pascal/Lazarus which scans a whole text file (in plain text, HTML, EPUB, or ODT formats) and ranks all used words according to frequency, performing a quantitative analysis of the text using Shannon-Weaver information statistic and Zipf power law function. It counts words, sentences, chars, spaces, and syllables. Also computes readability indexes (Gunning-Fog, Coleman-Liau, Automated Readability Index (ARI), SMOG grade, Flesch–Kincaid grade level and Flesch Reading Ease). -
18
-
19
GoodByeCatpcha
Solver ReCaptcha v2 Free
An async Python library to automate solving ReCAPTCHA v2 by images/audio using Mozilla's DeepSpeech, PocketSphinx, Microsoft Azure’s, Google Speech and Amazon's Transcribe Speech-to-Text API. Also image recognition to detect the object suggested in the captcha. Built with Pyppeteer for Chrome automation framework and similarities to Puppeteer, PyDub for easily converting MP3 files into WAV, aiohttp for async minimalistic web-server, and Python’s built-in AsyncIO for convenience. -
20
Zettel
Zettel allows taking notes from several references and organizing them
Zettel is a program for taking notes from bibliographic references. Instead of marking the text on paper and then going crazy looking for where the copy ended up, the notes are saved in a database, linked to the reference from where they were copied. Notes can be tagged and retrieved in several ways. Zettel é um programa para fichamento de referências bibliográficas. Ao invés de marcar o texto em papel e depois enlouquecer procurando onde a cópia foi parar, as notas são guardadas numa base de dados, vinculadas à referência de onde foram copiadas. Notas podem ser marcadas com tags, e recuperadas de várias formas. -
21
*nts* provides a simple format for using text files to store notes, a command line interface for viewing notes in a variety of convenient ways and a cross-platform, wx(python)-based GUI for creating and modifying notes as well as viewing them.
-
22
AI Researcher
An autonomous AI researcher
AI Researcher is an experimental open-source project that demonstrates how multiple AI agents can collaborate to conduct complex research tasks from start to finish with minimal human intervention. It orchestrates agents that can generate research questions, perform literature reviews, execute experiments, analyze results, and synthesize findings into structured outputs like reports or code. Each agent operates with clear roles — such as researcher, analyst, and summarizer — and they communicate through a task-management interface that ensures progress tracking and iterative refinement. The system emphasizes modularity, so teams can swap in new reasoning modules, data retrieval strategies, or domain knowledge bases depending on the research topic. Through self-supervised feedback loops, agents adjust their strategies based on prior outcomes, improving both the quality and relevance of results over time. -
23
AllenNLP
An open-source NLP research library, built on PyTorch
AllenNLP makes it easy to design and evaluate new deep learning models for nearly any NLP problem, along with the infrastructure to easily run them in the cloud or on your laptop. AllenNLP includes reference implementations of high quality models for both core NLP problems (e.g. semantic role labeling) and NLP applications (e.g. textual entailment). AllenNLP supports loading "plugins" dynamically. A plugin is just a Python package that provides custom registered classes or additional allennlp subcommands. There is ecosystem of open source plugins, some of which are maintained by the AllenNLP team here at AI2, and some of which are maintained by the broader community. AllenNLP will automatically find any official AI2-maintained plugins that you have installed, but for AllenNLP to find personal or third-party plugins you've installed, you also have to create either a local plugins file named .allennlp_plugins in the directory where you run the allennlp command. -
24
Auto File Selection
Detect all the "important" files from your computer.
The main aim of this project is to design and develop a mechanism that can find all the “important” files inside a computer. -
25
This project is intended to provide code to be used with MySQL and Python to create a database of major league baseball game events which are freely provided by the mlb.com Gameday application. Older version also support creating a retrosheet.org database but that is no longer supported. All major and minor league pitch location and game statistic data can be downloaded using BBOS. Installation Videos! Part 1: http://youtu.be/rnv2VLcG-eI Part 2: http://youtu.be/eFudbMWHNlQ Special thanks to Wells Oliver for the code for downloading Retrosheet files. And the Chadwick project for its Retrosheet tools. https://sourceforge.net/projects/chadwick/?source=recommended
