Open Source Windows Generative AI
Sort By:
Generative AI for Windows
View 1816 business solutionsBrowse free open source Generative AI and projects for Windows below. Use the toggles on the left to filter open source Generative AI by OS, license, language, programming language, and project status.
-
Gemini 3 and 200+ AI Models on One PlatformBuild generative AI apps with Vertex AI. Switch between models without switching platforms.
-
Fully Managed MySQL, PostgreSQL, and SQL ServerCloud SQL handles your database ops end to end, so you can focus on your app.
-
1
Diffusers
State-of-the-art diffusion models for image and audio generation
Diffusers is the go-to library for state-of-the-art pretrained diffusion models for generating images, audio, and even 3D structures of molecules. Whether you're looking for a simple inference solution or training your own diffusion models, Diffusers is a modular toolbox that supports both. Our library is designed with a focus on usability over performance, simple over easy, and customizability over abstractions. State-of-the-art diffusion pipelines that can be run in inference with just a few lines of code. Interchangeable noise schedulers for different diffusion speeds and output quality. Pretrained models that can be used as building blocks, and combined with schedulers, for creating your own end-to-end diffusion systems. We recommend installing Diffusers in a virtual environment from PyPi or Conda. For more details about installing PyTorch and Flax, please refer to their official documentation. -
2
Image Super-Resolution (ISR)
Super-scale your images and run experiments with Residual Dense
The goal of this project is to upscale and improve the quality of low-resolution images. This project contains Keras implementations of different Residual Dense Networks for Single Image Super-Resolution (ISR) as well as scripts to train these networks using content and adversarial loss components. Docker scripts and Google Colab notebooks are available to carry training and prediction. Also, we provide scripts to facilitate training on the cloud with AWS and Nvidia-docker with only a few commands. When training your own model, start with only PSNR loss (50+ epochs, depending on the dataset) and only then introduce GANS and feature loss. This can be controlled by the loss weights argument. The weights used to produce these images are available directly when creating the model object. ISR is compatible with Python 3.6 and is distributed under the Apache 2.0 license. -
3
NVIDIA NeMo
Toolkit for conversational AI
NVIDIA NeMo, part of the NVIDIA AI platform, is a toolkit for building new state-of-the-art conversational AI models. NeMo has separate collections for Automatic Speech Recognition (ASR), Natural Language Processing (NLP), and Text-to-Speech (TTS) models. Each collection consists of prebuilt modules that include everything needed to train on your data. Every module can easily be customized, extended, and composed to create new conversational AI model architectures. Conversational AI architectures are typically large and require a lot of data and compute for training. NeMo uses PyTorch Lightning for easy and performant multi-GPU/multi-node mixed-precision training. Supported models: Jasper, QuartzNet, CitriNet, Conformer-CTC, Conformer-Transducer, Squeezeformer-CTC, Squeezeformer-Transducer, ContextNet, LSTM-Transducer (RNNT), LSTM-CTC. NGC collection of pre-trained speech processing models. -
4
marqo
Tensor search for humans
A tensor-based search and analytics engine that seamlessly integrates with your applications, websites, and workflows. Marqo is a versatile and robust search and analytics engine that can be integrated into any website or application. Due to horizontal scalability, Marqo provides lightning-fast query times, even with millions of documents. Marqo helps you configure deep-learning models like CLIP to pull semantic meaning from images. It can seamlessly handle image-to-image, image-to-text and text-to-image search and analytics. Marqo adapts and stores your data in a fully schemaless manner. It combines tensor search with a query DSL that provides efficient pre-filtering. Tensor search allows you to go beyond keyword matching and search based on the meaning of text, images and other unstructured data. Be a part of the tribe and help us revolutionize the future of search. Whether you are a contributor, a user, or simply have questions about Marqo, we got your back. -
MongoDB Atlas runs apps anywhereMongoDB Atlas gives you the freedom to build and run modern applications anywhere—across AWS, Azure, and Google Cloud. With global availability in over 115 regions, Atlas lets you deploy close to your users, meet compliance needs, and scale with confidence across any geography.
-
5
DeepMozart
Audio generation using diffusion models
Audio generation using diffusion models in PyTorch. The code is based on the audio-diffusion-pytorch repository. -
6
Generative AI Docs
Documentation for Google's Gen AI site - including Gemini API & Gemma
Generative AI Docs is Google’s official documentation repository for Gemini, Vertex AI, and related generative AI APIs. It contains guides, API references, and examples for developers building applications using Google’s large language models, text-to-image models, embeddings, and multimodal capabilities. The repository includes markdown source files that power the Google AI developer documentation site, as well as sample code snippets in Python, JavaScript, and other languages that demonstrate how to use Google’s Generative AI SDKs and REST APIs effectively. -
7
Haystack
Haystack is an open source NLP framework to interact with your data
Apply the latest NLP technology to your own data with the use of Haystack's pipeline architecture. Implement production-ready semantic search, question answering, summarization and document ranking for a wide range of NLP applications. Evaluate components and fine-tune models. Ask questions in natural language and find granular answers in your documents using the latest QA models with the help of Haystack pipelines. Perform semantic search and retrieve ranked documents according to meaning, not just keywords! Make use of and compare the latest pre-trained transformer-based languages models like OpenAI’s GPT-3, BERT, RoBERTa, DPR, and more. Pick any Transformer model from Hugging Face's Model Hub, experiment, find the one that works. Use Haystack NLP components on top of Elasticsearch, OpenSearch, or plain SQL. Boost search performance with Pinecone, Milvus, FAISS, or Weaviate vector databases, and dense passage retrieval. -
8
Intelligent Java
Integrate with the latest language models, image generation and speech
Intelligent java (IntelliJava) is the ultimate tool to integrate with the latest language models and deep learning frameworks using java. The library provides an intuitive functions for sending input to models like ChatGPT and DALL·E, and receiving generated text, speech or images. With just a few lines of code, you can easily access the power of cutting-edge AI models to enhance your projects. Access ChatGPT, GPT3 to generate text and DALL·E to generate images. OpenAI is preferred for quality results without tuning. Generate text; Cohere allows you to generate a language model to suit your specific needs. Generate audio from text; Access DeepMind’s speech models. The only dependencies is GSON. Required to add manually when using IntelliJava jar. However, if you imported this repo through Maven, it will handle the dependencies. -
9
SentenceTransformers
Multilingual sentence & image embeddings with BERT
SentenceTransformers is a Python framework for state-of-the-art sentence, text and image embeddings. The initial work is described in our paper Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks. You can use this framework to compute sentence / text embeddings for more than 100 languages. These embeddings can then be compared e.g. with cosine-similarity to find sentences with a similar meaning. This can be useful for semantic textual similar, semantic search, or paraphrase mining. The framework is based on PyTorch and Transformers and offers a large collection of pre-trained models tuned for various tasks. Further, it is easy to fine-tune your own models. Our models are evaluated extensively and achieve state-of-the-art performance on various tasks. Further, the code is tuned to provide the highest possible speed. -
Full-stack observability with actually useful AI | Grafana CloudBuilt on open standards like Prometheus and OpenTelemetry, Grafana Cloud includes Kubernetes Monitoring, Application Observability, Incident Response, plus the AI-powered Grafana Assistant. Get started with our generous free tier today.
-
10
Diffusion WebUI Colab
Choose your diffusion models and spin up a WebUI on Colab in one click
The most simplistic Colab with most models included by default. Custom models can be added easily. Stable Diffusion 2.0 in testing phase. Choose your diffusion models and spin up a WebUI on Colab in one click. Share your generations in our mastodon server - (This is hosted by a third party. I am not associated with the instance in any way.) The instructions are on the Colab. -
11
FID score for PyTorch
Compute FID scores with PyTorch
This is a port of the official implementation of Fréchet Inception Distance to PyTorch. FID is a measure of similarity between two datasets of images. It was shown to correlate well with human judgement of visual quality and is most often used to evaluate the quality of samples of Generative Adversarial Networks. FID is calculated by computing the Fréchet distance between two Gaussians fitted to feature representations of the Inception network. The weights and the model are exactly the same as in the official Tensorflow implementation, and were tested to give very similar results (e.g. .08 absolute error and 0.0009 relative error on LSUN, using ProGAN generated images). However, due to differences in the image interpolation implementation and library backends, FID results still differ slightly from the original implementation. In difference to the official implementation, you can choose to use a different feature layer of the Inception network instead of the default pool3 layer. -
12
GPT-NeoX
Implementation of model parallel autoregressive transformers on GPUs
This repository records EleutherAI's library for training large-scale language models on GPUs. Our current framework is based on NVIDIA's Megatron Language Model and has been augmented with techniques from DeepSpeed as well as some novel optimizations. We aim to make this repo a centralized and accessible place to gather techniques for training large-scale autoregressive language models, and accelerate research into large-scale training. For those looking for a TPU-centric codebase, we recommend Mesh Transformer JAX. If you are not looking to train models with billions of parameters from scratch, this is likely the wrong library to use. For generic inference needs, we recommend you use the Hugging Face transformers library instead which supports GPT-NeoX models. -
13
Generative AI
Sample code and notebooks for Generative AI on Google Cloud
Generative AI is a comprehensive collection of code samples, notebooks, and demo applications designed to help developers build generative-AI workflows on the Vertex AI platform. It spans multiple modalities—text, image, audio, search (RAG/grounding) and more—showing how to integrate foundation models like the Gemini family into cloud projects. The README emphasises getting started with prompts, datasets, environments and sample apps, making it ideal for both experimentation and production-ready usage. The repository architecture is organised into folders like gemini/, search/, vision/, audio/, and rag-grounding/, which helps developers locate use cases by modality. It is licensed under Apache-2.0, opensourced and maintained by Google, meaning it's designed with enterprise-grade practices in mind. Overall, it serves as a practical entry point and reference library for building real-world generative AI systems on Google Cloud. -
14
Swirl
Swirl queries any number of data sources with APIs
Swirl queries any number of data sources with APIs and uses spaCy and NLTK to re-rank the unified results without extracting and indexing anything! Includes zero-code configs for Apache Solr, ChatGPT, Elastic Search, OpenSearch, PostgreSQL, Google BigQuery, RequestsGet, Google PSE, NLResearch.com, Miro & more! SWIRL adapts and distributes queries to anything with a search API - search engines, databases, noSQL engines, cloud/SaaS services etc - and uses AI (Large Language Models) to re-rank the unified results without extracting and indexing anything. It's intended for use by developers and data scientists who want to solve multi-silo search problems from enterprise search to new monitoring & alerting solutions that push information to users continuously. Built on the Python/Django/RabbitMQ stack, SWIRL includes connectors to Apache Solr, ChatGPT, Elastic, OpenSearch | PostgreSQL, Google BigQuery plus generic HTTP/GET/JSON with configurations for premium services. -
15
BertViz
BertViz: Visualize Attention in NLP Models (BERT, GPT2, BART, etc.)
BertViz is an interactive tool for visualizing attention in Transformer language models such as BERT, GPT2, or T5. It can be run inside a Jupyter or Colab notebook through a simple Python API that supports most Huggingface models. BertViz extends the Tensor2Tensor visualization tool by Llion Jones, providing multiple views that each offer a unique lens into the attention mechanism. The head view visualizes attention for one or more attention heads in the same layer. It is based on the excellent Tensor2Tensor visualization tool. The model view shows a bird's-eye view of attention across all layers and heads. The neuron view visualizes individual neurons in the query and key vectors and shows how they are used to compute attention. -
16
Bottery
A conversational agent prototyping platform
Bottery is a prototyping environment and mini-language for designing conversational agents as explicit state machines rather than opaque bundles of code. It encourages authors to model intents, dialog turns, and transitions in a compact, declarative form that’s easy to visualize and simulate. A built-in simulator lets you “play” the conversation, test edge cases, and tweak flows without wiring up back-end services first. Because behavior is defined in a DSL, teams can version the conversation design alongside code and keep product, research, and engineering aligned. The repo includes examples of slots, branching, and side effects, showing how to express common patterns like confirmations, fallbacks, and small talk. Bottery’s value is in reducing the gap between whiteboard diagrams and a running prototype you can test with real users. -
17
ChatGPT-Reviewer
Automated pull requests reviewing and issues triaging with ChatGPT
Automated pull requests reviewing and issues triaging with ChatGPT. Create an OpenAI API key here, and then set the key as an action secret in your repository named OPENAI_API_KEY. The ChatGPT reviewer PRs are also getting reviewed by ChatGPT, refer the pull requests for the sample review comments. In order to protect public repositories for malicious users, Github runs all pull request workflows raised from repository forks with a read-only token and no access to secrets. -
18
CodiumAI PR-Agent
AI-Powered tool for automated pull request analysis
CodiumAI PR-Agent is an open-source tool aiming to help developers review pull requests faster and more efficiently. It automatically analyzes the pull request and can provide several types of commands. See the Usage Guide for instructions how to run the different tools from CLI, online usage, Or by automatically triggering them when a new PR is opened. You can try GPT-4 powered PR-Agent, on your public GitHub repository, instantly. Just mention @CodiumAI-Agent and add the desired command in any PR comment. The agent will generate a response based on your command. -
19
Edward
A probabilistic programming language in TensorFlow
A library for probabilistic modeling, inference, and criticism. Edward is a Python library for probabilistic modeling, inference, and criticism. It is a testbed for fast experimentation and research with probabilistic models, ranging from classical hierarchical models on small data sets to complex deep probabilistic models on large data sets. Edward fuses three fields, Bayesian statistics and machine learning, deep learning, and probabilistic programming. Edward is built on TensorFlow. It enables features such as computational graphs, distributed training, CPU/GPU integration, automatic differentiation, and visualization with TensorBoard. Expectation-Maximization, pseudo-marginal and ABC methods, and message passing algorithms. -
20
GPT2 for Multiple Languages
GPT2 for Multiple Languages, including pretrained models
With just 2 clicks (not including Colab auth process), the 1.5B pretrained Chinese model demo is ready to go. The contents in this repository are for academic research purpose, and we do not provide any conclusive remarks. Research supported with Cloud TPUs from Google's TensorFlow Research Cloud (TFRC) Simplifed GPT2 train scripts（based on Grover, supporting TPUs). Ported bert tokenizer, multilingual corpus compatible. 1.5B GPT2 pretrained Chinese model (~15G corpus, 10w steps). Batteries-included Colab demo. 1.5B GPT2 pretrained Chinese model (~30G corpus, 22w steps). -
21
Generative AI JS
This SDK is now deprecated, use the new unified Google GenAI SDK
deprecated-generative-ai-js is a JavaScript/TypeScript client and example suite for interacting with Gemini generative APIs in web and Node.js environments. Though marked deprecated (likely superseded by newer SDKs), the repo shows how to wrap HTTP/WS endpoints, manage streaming responses, and interoperate with browser UI or server logic. The examples include chat widgets, prompt pipelines, and generalized inference utilities. It also deals with streaming cancellation, retries, backoff logic, and message chunk assembly to help developers handle real-world use. Because it’s JavaScript, the repo supports both ESM and CommonJS contexts, making it versatile in backend and frontend setups. The deprecation label reflects that newer or official SDKs may have replaced it, but many of its patterns still serve as a useful reference to understand how streaming, chunking, and prompt logic can be implemented by hand in JS. -
22
Generative AI Swift
This SDK is now deprecated, use the unified Firebase SDK
deprecated-generative-ai-swift is a Swift client and example scaffold for building generative AI apps using the Gemini models. Although marked “deprecated”, the repo demonstrates how to integrate Gemini inference into iOS and macOS apps via Swift APIs, providing boilerplate for prompt dispatching, streaming responses, UI integration, and error handling. It includes a sample app that showcases a chat interface, where users send messages and receive responses streamed in real time, with UI updates as tokens arrive. The code also handles request queuing, cancellation, and retry logic, giving developers a realistic foundation rather than a minimalist “hello world.” Despite its deprecated label, the repo remains valuable for developers who want to see how a native Swift integration might be structured before migrating to newer SDKs. Maintainability is emphasized: modular layers separate networking, prompt handling, and UI logic, making adaptation easier when switching to updated APIs. -
23
MMGeneration
MMGeneration is a powerful toolkit for generative models
MMGeneration has been merged in MMEditing. And we have supported new-generation tasks and models. MMGeneration is a powerful toolkit for generative models, especially for GANs now. It is based on PyTorch and MMCV. The master branch works with PyTorch 1.5+. We currently support training on Unconditional GANs, Internal GANs, and Image Translation Models. Support for conditional models will come soon. A plentiful toolkit containing multiple applications in GANs is provided to users. GAN interpolation, GAN projection, and GAN manipulations are integrated into our framework. It's time to play with your GANs! For the highly dynamic training in generative models, we adopt a new way to train dynamic models with MMDDP. A new design for complex loss modules is proposed for customizing the links between modules, which can achieve flexible combinations among different modules. Conditional GANs have been supported in our toolkit. More methods and pre-trained weights will come soon. -
24
NiftyNet
An open-source convolutional neural networks platform for research
An open-source convolutional neural networks platform for medical image analysis and image-guided therapy. NiftyNet is a TensorFlow-based open-source convolutional neural networks (CNNs) platform for research in medical image analysis and image-guided therapy. NiftyNet’s modular structure is designed for sharing networks and pre-trained models. Using this modular structure you can get started with established pre-trained networks using built-in tools. Adapt existing networks to your imaging data. Quickly build new solutions to your own image analysis problems. NiftyNet currently supports medical image segmentation and generative adversarial networks. NiftyNet is not intended for clinical use. -
25
PaddleGAN
PaddlePaddle GAN library, including lots of interesting applications
PaddlePaddle GAN library, including lots of interesting applications like First-Order motion transfer, Wav2Lip, picture repair, image editing, photo2cartoon, image style transfer, GPEN, and so on. PaddleGAN provides developers with high-performance implementation of classic and SOTA Generative Adversarial Networks, and supports developers to quickly build, train and deploy GANs for academic, entertainment, and industrial usage. GAN-Generative Adversarial Network, was praised by "the Father of Convolutional Networks" Yann LeCun (Yang Likun) as [One of the most interesting ideas in the field of computer science in the past decade]. It's the one research area in deep learning that AI researchers are most concerned about.
