Best open source AI tools

Captum is an open-source model interpretability and explainability library for PyTorch. It provides state-of-the-art algorithms to help developers and researchers understand which features influence a model's predictions. Supporting multi-modal data like text, vision, and more, Captum makes it easy to debug models, improve transparency, and benchmark new interpretability techniques within the PyTorch ecosystem.

AIaC by Firefly is a free, open-source CLI tool that uses AI to generate Infrastructure-as-Code (IaC) from natural language prompts. It helps developers and DevOps engineers quickly create configuration code for various cloud platforms and IaC frameworks like Terraform, Pulumi, and CloudFormation, significantly speeding up development and reducing syntax errors.

Magi, by Sand.ai, is a revolutionary AI video platform featuring the world's first autoregressive video model, Magi-1. It transforms images or text into high-quality videos, offers infinite video extension capabilities, and creates lifelike talking avatars from a single photo. With an open-source model, Magi provides creators with unparalleled control, speed, and creative freedom for filmmaking, marketing, and content creation.

LangUI is an open-source Tailwind CSS component library offering over 60 free, ready-to-use components specifically designed for AI and GPT projects. It helps developers quickly build beautiful, responsive user interfaces for their AI applications, allowing them to focus on core functionality.

Janus Pro AI is a powerful open-source multimodal model developed by Deepseek. It unifies image understanding and text-to-image generation within a single framework. Outperforming models like DALL-E 3 in benchmarks, it offers 1B and 7B parameter versions under an MIT license, making it ideal for both research and unrestricted commercial use. It's designed for high performance, flexibility, and cost-effective scalability.

nv_tlabs is NVIDIA's research hub, showcasing a portfolio of cutting-edge AI projects. It provides access to pioneering research papers, interactive demos, and open-source code in fields like generative AI, computer vision, and neural graphics, targeting researchers and developers.

ChatGPT EVO is a powerful suite of open-source browser extensions and userscripts designed to augment your web experience. It seamlessly integrates AI functionalities from models like ChatGPT into your daily browsing, including search engines, e-commerce sites, and the browser's address bar, boosting productivity and providing instant information.

An AI-powered tool that analyzes your GitHub profile and starred repositories to generate insightful summaries and witty roasts. Discover what your coding interests truly reveal about you in a fun, shareable format.

GenWorlds is an open-source, event-based framework for building and coordinating complex multi-agent AI systems. It allows developers to create customizable worlds where multiple AI agents, each with unique personalities, memories, and cognitive processes, can collaborate to perform complex tasks. It's built on LangChain and uses Qdrant for long-term memory.

Dyad is a free, local, and open-source AI app builder that allows you to create full-stack applications without coding. It runs directly on your machine, ensuring privacy and no vendor lock-in, while supporting various AI models like GPT-4, Gemini, and local models via Ollama.

xMem is a hybrid memory orchestrator for LLMs, designed to give AI applications persistent memory. It combines long-term knowledge from vector databases with real-time session context, enabling LLMs to remember past interactions and deliver smarter, more relevant responses without losing context between sessions.

Galactica is a large language model from Meta AI, specifically trained on over 48 million scientific papers, textbooks, and reference materials. It's designed to assist researchers by organizing scientific knowledge, suggesting citations, answering complex questions, writing scientific code, and explaining mathematical formulas. Although its public demo is discontinued, the open-source model remains available for the research community to advance scientific discovery.

The official open-source hub for Google on GitHub. It hosts a vast collection of over 2,700 public repositories, including libraries, developer tools, frameworks, and style guides. A critical resource for developers working with Java, C++, Python, Android, web technologies, and more, offering production-tested code and fostering community collaboration.

An AI-powered Figma plugin that automatically renames all your design layers in a single click. It intelligently preserves component names, supports custom naming models, and is completely free and open-source, saving designers countless hours of tedious work.

An open-source Google research project that explores the role of machine learning in the creative process. Magenta provides AI-powered tools and plugins, like The Infinite Crate, for real-time music generation directly within Digital Audio Workstations (DAWs) using text prompts.

Fig was a popular open-source tool that added IDE-style visual autocomplete to the command line. It has been acquired by AWS and is now sunset, with users encouraged to migrate to its successor, Amazon Q for command line, which is free for individuals.

TensorFlow is an end-to-end open-source platform for machine learning developed by Google. It provides a comprehensive, flexible ecosystem of tools, libraries, and community resources that lets researchers and developers build and deploy ML-powered applications. From beginners to experts, TensorFlow offers intuitive high-level APIs for easy model building and powerful low-level APIs for advanced research, enabling deployment across servers, edge devices, and browsers.

FollowFox offers Distillery, an open-source AI text-to-image generator. It provides advanced control with over 25 parameters, 1-image training for custom avatars, and a commitment to accessibility with 10 free daily generations. Ideal for artists, developers, and hobbyists seeking creative freedom and transparency.

Clack is an AI-powered writing productivity tool that effortlessly tracks your writing progress from Google Docs and Notion. It visualizes your activity on a calendar heatmap, offers a Notion-style editor with AI assistance, and helps you build consistent writing habits.

CrewAI is a powerful multi-agent platform for building and orchestrating collaborative AI agent workflows. It enables developers to create "crews" of specialized AI agents that work together to automate complex tasks. With its open-source framework, no-code UI Studio, and "Flows" feature for structured automation, it streamlines development from planning to deployment and monitoring, integrating with any LLM and cloud provider.

AFFiNE is an all-in-one, AI-powered KnowledgeOS that seamlessly merges documents, whiteboards, and databases. It's a privacy-focused, open-source workspace designed to enhance productivity by allowing you to write, draw, and plan in a single, unified environment, eliminating the need to switch between multiple applications.

GitHub Next is GitHub's research and innovation lab, dedicated to exploring the future of software development. It pioneers new tools and technologies, with a strong focus on AI, to enhance developer productivity, collaboration, and overall experience.

Supabase is an open-source Firebase alternative, providing a complete backend solution built on Postgres. It offers a suite of tools including a database, authentication, instant APIs, edge functions, real-time subscriptions, storage, and vector embeddings to accelerate application development from prototype to production.

Rerun is an open-source data stack for Physical AI, providing powerful logging and visualization tools for multimodal, time-series data. Designed for robotics, computer vision, and spatial computing, it helps developers understand and debug complex systems with SDKs for Python, Rust, and C++.