Ollama
Run and manage compatible language models locally with a CLI and API.
For: Try local models on a personal computer and connect them to an application.
Discover open source AI projects, from inspiration to practice.
Curated projects for model training, applications, agents, and development
33 projects
Run and manage compatible language models locally with a CLI and API.
For: Try local models on a personal computer and connect them to an application.
A framework for durable, stateful agent workflows and multi-step applications.
For: Build stateful agent workflows with review and multiple execution steps.
A node-based generative AI workflow tool for composing models and steps.
For: Build an image-generation workflow and check model and hardware requirements.
C/C++ inference for language models across hardware and quantization options.
For: Explore lightweight deployment or run compatible GGUF models.
An open-source engine for document parsing and retrieval-augmented generation.
For: Prototype an internal knowledge base and check retrieval quality.
Inference examples and tools for OpenAI’s open-weight model series.
For: Evaluate hardware needs and experiment with open-weight models locally.
Tools for loading, training and running text, vision, audio and other models.
For: Prototype tasks with pretrained models or learn model APIs.
Build AI applications visually by connecting models, data and components.
For: Prototype a knowledge-base assistant with a visual workflow.
General-purpose speech recognition models and code for transcription.
For: Transcribe authorized recordings and review the output.
A high-throughput inference and serving engine for large language models.
For: Serve models to a team and benchmark throughput on your hardware.
Optimization tools for training and serving large deep-learning models.
For: Experiment with distributed training or model resource optimization.
Parameter-efficient fine-tuning tools for adapting supported models.
For: Adapt a model with techniques such as LoRA using authorized data.
A toolkit for language-model post-training and alignment.
For: Experiment with supervised fine-tuning or preference optimization.
An AI workspace for document Q&A with support for different model providers.
For: Create a document Q&A workspace and check its access controls.
A document retrieval and question-answering interface with configurable pipelines.
For: Build a retrieval demo for documents you are authorized to use.
A multi-agent development framework for coordinating roles and tasks.
For: Explore coordinating multiple roles on a complex task.
A framework for building model-powered applications and tool workflows.
For: Combine models, retrieval and tools into an application prototype.
Microsoft’s SDK for combining models, plugins and agent capabilities.
For: Integrate models and callable capabilities into an application.
A framework for organizing agents and tasks that work together.
For: Experiment with research, writing and review roles.
A lightweight library for agents that use tools to complete tasks.
For: Build a small tool-using agent with restricted execution permissions.
A Python SDK for building agent applications with tools and handoffs.
For: Prototype an assistant that uses tools and hands tasks off.
A developer tool for agents that operate browsers and perform web tasks.
For: Test browser forms and workflows with test accounts and data.
Data tools for language-model applications, commonly used for retrieval.
For: Index authorized documents and evaluate answer citations.
An open-source framework for retrieval, Q&A and agent pipelines.
For: Compose document processing, retrieval and generation pipelines.
A RAG framework exploring graph-based indexing and retrieval.
For: Compare graph-based retrieval with a basic vector-search approach.
A vector database and search engine.
For: Store document vectors for semantic search or knowledge retrieval.
A library for diffusion-model inference and training.
For: Explore text-to-image pipelines or adapt diffusion models.
A Whisper inference implementation based on CTranslate2.
For: Benchmark transcription speed and accuracy on your hardware.
A toolkit for text recognition and document processing tasks.
For: Extract text from receipts or documents and review the results.
An open-source AI coding assistant integrating development tools and models.
For: Explore code explanation, editing and review in an editor.
An open-source coding agent for development workflows.
For: Try code changes in a reversible project and review commands.
An AI pair-programming tool for editing local code in a terminal.
For: Make a small change in a local repository and review the diff.
A tool that converts various document formats to Markdown for further processing.
For: Convert authorized documents to Markdown for retrieval or text processing.
Licenses shown apply to the code repositories. Model weights, datasets, and cloud services may have different terms; check each project’s official documentation before use.