Open Source Generative AI Projects
Generative AI uses models to create new content, such as text, images, audio, video, and code, in response to prompts or other inputs. It can help with tasks that involve drafting, summarizing, search, design, and content transformation. Building these systems also raises practical challenges, including model selection, data handling, evaluation, inference costs, and integration with existing applications.
Open source tools in this area include model training and fine-tuning frameworks, inference servers, application libraries, evaluation and monitoring utilities, and creative interfaces. When choosing a tool, consider its license, maintenance activity, documentation, hardware and software requirements, supported models, and compatibility with your workflow. These tools are useful to developers, researchers, creators, and organizations that want to build, adapt, or operate generative AI systems with greater control.
76 repositories · updated October 16, 2026
image-blaster: Turn Images into 3D Environments and Assets
image-blaster uses Claude skills and external generation services to turn an image into a 3D environment, object meshes, and sound effects. It is designed to quickly create starting points for game, 3D, and web projects.
QuietHealth 17/100
verl-omni: Train Multimodal Generative Models with RL
VeRL-Omni is a Python framework for reinforcement-learning post-training of diffusion, unified multimodal, and omni-modality models. It builds on verl and targets teams that need distributed training, multimodal rollouts, and configurable reward workflows.
NewHealth 68/100
guaardvark: Run a Local AI Studio on Your GPU
Guaardvark combines local AI media generation, coding agents, RAG, and automation in a self-hosted studio. It suits people who want these workflows on their own hardware and can meet its Python and GPU requirements.
ActiveHealth 62/100
ArcReel: Turn Stories and Scripts Into Editable Videos
ArcReel is a self-hosted workspace for turning novels, scripts, or product materials into storyboards and short videos. It organizes AI-generated assets and production steps, tracks costs, and can export a CapCut-compatible workflow as a Jianying draft for further editing.
ActiveHealth 72/100
dify-official-plugins: Extend Dify with Models and Tools
Official plugins for Dify add models, tools, agent strategies, and HTTP webhook extensions to the platform. Use this repository when building or managing Dify applications that need these integrations.
ActiveHealth 77/100
json-render: Build Guardrailed Interfaces from AI-Generated JSON
json-render turns AI-generated JSON into interfaces using components and actions you define, so model output stays within your application's boundaries. It supports multiple UI frameworks and targets beyond web interfaces, including mobile, email, PDFs, video, and 3D.
ActiveHealth 51/100
Speech: Build and Adapt Speech AI Models
NVIDIA NeMo Speech is a Python framework for developing and deploying speech models, including speech recognition, text-to-speech, and speech language models. It suits teams that want to adapt pretrained models or build speech systems with PyTorch.
ActiveHealth 68/100
rag-zero-to-hero-guide: Learn Retrieval-Augmented Generation
A learning guide to retrieval-augmented generation, from core concepts to evaluation and advanced approaches. It combines explanations, Jupyter notebook implementations, tool references, and survey papers for learners building RAG knowledge.
InactiveHealth 17/100
mergoo: Combine Fine-Tuned LLM Experts into Routed Models
Mergoo combines fine-tuned language models or LoRA adapters into routed expert models, then supports training the resulting model. It is aimed at teams that want one model to draw on specialized experts rather than use them separately.
InactiveHealth 26/100
lamini: Access the Lamini API from Python
Lamini is a Python client and SDK for working with Lamini's generative AI API. It suits developers who want to connect Python applications to Lamini without building API requests from scratch.
InactiveHealth 31/100
xTuring: Fine-Tune and Run Personalized LLMs
xTuring is a Python library for preparing data, fine-tuning, evaluating, and running open-source language models locally or in a private cloud. It is aimed at developers who want model customization through a high-level API and methods such as LoRA and quantization.
ActiveHealth 42/100
torchtune: Fine-Tune and Post-Train Large Language Models
torchtune is a PyTorch library for configuring and running LLM post-training workflows, from supervised fine-tuning to preference optimization. It is aimed at developers who want editable recipes and model-specific configs, but is no longer actively maintained.
QuietHealth 39/100