Open-source AI tools

AudioCraft is a library of generative audio and AI music tools and online demos launched by Meta AI, integrating capabilities such as MusicGen text-to-music, AudioGen text-to-sound effects, and EnCodec neural audio compression. You can quickly generate different styles of soundtracks, ambient sounds, and realistic sound effects with a single prompt, and control the rhythm, atmosphere, and duration through prompts, which is suitable for short video soundtracks, game sound prototypes, advertising ambient sounds, and creative inspiration verification. AudioCraft also provides open-source code and models for developers to deploy on-premises, integrate into workflows, and develop reactively.

XiaoZhi AI is an open-source AI hardware project and console service for developers and hardware vendors, focusing on low-latency voice conversations and scalable agent capabilities. Based on the Model Context Protocol (MCP), Xiaozhi AI connects speech recognition, intent understanding, memory and tool calling in series, and adapts to a variety of common chips and development boards, making it convenient and quick to make products such as emotional companionship, smart home control, desktop voice assistants, and in-vehicle robots. Through the Xiaozhi AI console, users can create agents, configure roles and behaviors, and complete device binding and management, making Xiaozhi AI more worry-free from prototyping to mass production access.

Deta Surf is an AI browser and smart notebook that blends browsing and note-taking, allowing web pages, documents, and ideas to flow naturally in one place. Deta Surf supports web search, PDF and YouTube embedded reading and questioning, automatically generates key point summaries and in-depth analysis, and keeps traceable citation links in your notes. Collect sites, images, and files with Notebooks, organize information with Vertical Tabs, and create interactive apps and visualizations with Surflets. Deta Surf is open-source and local-first, with data stored on the device and importable for import and export; Support for large models or local LLMs of your choice, using your own keys to meet the needs of privacy, security, and efficient research.

Helium Browser is a privacy-first, open-source browser that is safe to use with unbiased ad blocking and tracking protection enabled by default, blocking third-party cookies, and minimizing fingerprinting. Based on Chromium, it is compatible with all extensions and anonymously access the extension store; The interface is light, fast to start, and supports Split View split screen and native !bangs to quickly and directly access websites, and can also be used offline. With no built-in cloud sync and password management, Helium Browser offers self-hosted options and fast security updates for users and teams that value efficiency, control, and data sovereignty.

Browser Operator is an open-source, privacy-friendly AI browser for research, analysis, and automation scenarios. Browser Operator is based on the Chromium architecture, supports local AI inference and cloud large model collaboration, and provides AI Agent Studio to build AI agent processes, automatically capture web page information, extract key points, and generate summaries, and can complete forms, price comparisons, and lead sorting across multiple sites. Browser Operator has built-in research assistants, smart shopping and talent search templates, lowering the threshold for construction; The whole process is transparent and auditable, and the data is locally controllable. Currently in alpha with priority support for macOS, it's suitable for individuals and teams that value privacy, security, and efficiency, elevating their daily browsing into orchestrated AI browser workflows.

BrowserOS is an open-source, privacy-first AI browser that focuses on "turning words into actions". BrowserOS uses local AI agents to understand natural language commands and automatically complete web page operations such as clicking, typing, and jumping, making it suitable for high-frequency research and form process automation. It supports Split View to put models such as ChatGPT, Claude, and Gemini in sidebar conversations, making it more efficient to ask questions while reading; Compatible with the Chromium ecosystem, commonly used Chrome extensions and data can be seamlessly migrated. BrowserOS provides MCP server integration, which can connect to Gmail, Calendar, Docs, Sheets, and Notion with one click, upgrading the browser to a work hub. It also supports local models such as Ollama and LM Studio or comes with its own API key to ensure that the data is locally controllable. BrowserOS covers macOS, Windows, and Linux across platforms, and has built-in semantic retrieval and highlighting to comprehensively improve information retrieval and web office efficiency.

Nook Browser is an open-source, privacy-focused, and modern browser. Nook Browser takes local priority as the core, and the data is saved on your device, and you can export and delete it with one click, reducing the risk of synchronization and leakage. It organizes work and interests through Rooms/Spaces, avoids overwhelming tabs, and offers Peek Quick Preview and Split View split screen for efficient multitasking. Nook Browser pursues a simple, less intrusive online experience, with an open roadmap and community co-creation mechanism, making it easy to scale and customize. For users and teams that value privacy, efficiency, and sustainable evolution, Nook Browser is the ideal browser choice for lightweight and scalability.

Zen Browser is an open-source browser that emphasizes focus and efficiency, advocating for a "quieter internet." Zen Browser prioritizes privacy and reduces distractions and tracking, making it suitable for daily online and web work that requires stability and security. It offers features like Workspaces, Compact Mode, Glance Quick Switching, and Split View for clearer multitasking management. Zen Browser is continuously iterated by the community, with a beautiful interface, fast performance, and strong scalability, which can meet the high-frequency usage scenarios of developers and content creators. If you value privacy, efficiency, and comfort, Zen Browser is the browser choice that balances aesthetics and functionality.

Ladybird is a new non-profit-led, independent web browser and web engine that emphasizes web standards first and built from the ground up. Ladybird aims to be cross-platform, focusing on Linux and macOS at this stage, pursuing performance, stability, and security, without introducing other browser kernel code, nor commercializing it through default search shares or tokens. The project is open source and welcoming developers, collaborating with the community through GitHub to advance progress, with an alpha version planned for early users in 2026, suitable for users and development teams who value openness, transparency, and control.

Scira AI is an open-source AI search engine and Perplexity alternative that focuses on "real-time, verifiable, and self-hostable." It combines RAG (Retrieval-Augmented Generation) with Search Grounding to return answers with source citations, making it suitable for research, learning, and knowledge retrieval scenarios. Scira AI supports multi-model switching, covering Grok, Claude, DeepSeek, Qwen, OpenAI, etc., providing PDF analysis, history and voice reading capabilities, and realizes topic monitoring and timing tracking through Scira Lookout. Q: What is Scira AI? A: Scira AI is an open-source AI search tool that provides real-time search with trusted citations. As an alternative to Perplexity, Scira AI combines speed, transparency, and control, and can be used directly on the official website or self-built to meet the AI search needs of individuals and teams.

Cloudflare VibeSDK is an open-source reference implementation for AI programming that helps teams quickly build and deploy AI applications on the Cloudflare Developer Platform. It takes the core path of "from natural language to runnable applications", combining Workers, Pages, KV, Durable Objects, Queues, R2, Vectorize, and AI Gateway to cover full-link scenarios of generation, storage, inference, and observation. Brief definition: Cloudflare VibeSDK is an all-in-one sample platform for AI application development and delivery on the Cloudflare edge network. Common use cases include intelligent customer service, retrieval-augmented generation, and full-stack chat applications for teams looking to validate and scale their AI products with low operational costs.

Genmo Mochi 1 is an open-source AI video generation tool for creators and teams, supporting text-to-video, image-to-video, and per-shot control, highlighting physical consistency, character coherence, and high-prompt alignment. Genmo Mochi 1 offers both cloud playground and private deployment, making it easy to quickly produce in scenarios such as advertising storyboards, product demonstrations, educational animations, and social media videos. As an "open-source AI video generation model", Genmo Mochi 1 is easy to use and scalable, making it suitable for integrating AI video generation into existing creative processes and production systems.

21st.dev is a UI component discovery and management platform built for front-end design and development teams, integrating AI intelligent agents to retrieve and import high-quality components in mainstream IDEs with one click. Users can browse, share, replicate, and customize open-source components with real-time preview and version control to help unify team visual style and development specifications. The platform provides comprehensive API interfaces and Magic MCP functions, supporting batch generation, parametric customization, and rapid prototype iteration, seamlessly connecting design and code processes. The user-friendly interface and community-driven ecosystem allow design engineers and "vibe coders" to unleash their creativity in collaboration, improving project development efficiency and consistency. With no on-premises deployment required, cloud collaboration and exportable seed projects allow teams to easily expand and maintain component libraries and shorten project time-to-live.

OpenVoiceOS (OVOS) is a community-driven, open-source voice AI platform designed to create custom voice-controlled interfaces for various devices. The platform emphasizes privacy and security, allowing users to process voice data locally and avoid sending sensitive information to the cloud, enhancing data protection. OVOS supports a wide range of hardware platforms, including Raspberry Pi, Mycroft devices, and Linux desktops and laptops, for embedded systems and low-profile devices. Its modular architecture includes components such as ovos-core, ovos-listener, and ovos-messagebus, and supports plug-in speech recognition (STT) and text-to-speech (TTS) engines, allowing users to choose the appropriate plug-in according to their needs. OVOS also provides a wealth of developer tools and documentation to facilitate developers to create and deploy custom voice applications. As a continuation of the Mycroft project, OpenVoiceOS is committed to providing a voice assistant solution that is transparent, customizable, and respects user privacy.

n8n is an open-source workflow automation platform that combines visual build with code flexibility and supports over 500 application integrations. Users can create multi-step automated processes through drag-and-drop interfaces or JavaScript/Python scripts, suitable for IT operations, security response, sales analysis, and other scenarios. n8n has built-in AI capabilities that support the construction of LLM-based agent systems for natural language processing, data extraction, and automated decision-making. The platform offers self-hosted and cloud deployment options, ensuring data security and compliance. With its strong scalability and active community support, n8n is an ideal choice for technology teams to achieve efficient automation.

DorkGPT is an AI-powered Google Dork query generator tool designed to help users efficiently construct advanced search statements to reveal hidden web page data and potential security vulnerabilities. Through natural language processing technology, users only need to input simple query requirements, and DorkGPT can automatically generate accurate Google Dork statements, suitable for various scenarios such as cybersecurity audits, open-source intelligence (OSINT) research, and academic data retrieval. With an emphasis on responsible use, the platform encourages users to explore information within a legal and ethical framework, making it an ideal aid for security researchers, journalists, and data analysts.

Morphic is a fully open-source AI search engine that combines a generative user interface (Generative UI) to provide natural language Q&A, web summarization, video search, and more. The platform supports multiple AI models (such as OpenAI, Anthropic, Google Generative AI, Groq, Ollama, etc.) and integrates search engines such as Tavily, SearXNG, and Exa to cater to diverse search needs. Morphic offers features such as URL-specific search, video search, model switching, search history, and result sharing, making it suitable for developers, researchers, and teams with high data privacy requirements. Users can deploy the local version via Docker or Vercel, or use the official hosting service to easily integrate into the browser as the default search engine. Morphic stands out as an ideal platform for exploring AI search and generative interactions with its flexibility and scalability.

OSS Insight is an open-source data analysis platform launched by PingCAP, focusing on visual insights into the GitHub open-source ecosystem. Its core feature, GitHub Data Explorer, allows users to ask questions in natural language, automatically generate SQL queries, and analyze over 8 billion GitHub event data in real-time. The platform is based on TiDB Cloud and Chat2Query technology stack, which supports users without programming experience for data exploration. The main functions include: developer behavior analysis, repository trend comparison, technical field rankings, geographical distribution of contributors, etc. Users can also view trends in open source projects in popular technology fields such as Web3, AI, and databases through the Collections module. OSS Insight provides 2D and 3D visualization interfaces for developers, data analysts, and researchers to gain insights into the open source community.

LlamaCoder is an open-source AI programming platform launched by Together AI, designed to quickly generate complete web applications from natural language prompts. The platform is built on Meta's Llama 3.1 405B model, which supports the generation of React applications and components, and users only need to describe the application requirements, and LlamaCoder can generate corresponding code. The platform offers both online and on-premises deployments, catering to the needs of different developers. LlamaCoder integrates technology stacks such as Sandpack, Next.js, Tailwind CSS, etc., supporting real-time preview and one-click deployment, suitable for rapid prototyping and product iteration. Its open-source nature and active community support make it an ideal tool for developers exploring AI programming.

Dreamlook is an online AI image generation and model fine-tuning platform for creative teams and developers. By optimizing Stable Diffusion, users can complete model training in minutes without on-premise deployment, increasing the training speed by 2.5 times and supporting thousands of concurrent runs per day. The platform provides flexible API interfaces and multi-level packages to meet the needs of projects of different sizes, and automatically clears all uploaded data within 48 hours after the end of training to ensure privacy and security. Rich LoRA export and high-resolution output capabilities help users quickly generate professional-grade customized visual content.

AIimag.es is an open-source offline AI drawing software based on Stable Diffusion, which supports Windows system installation, and users only need to enter Chinese or English prompts to quickly generate high-definition pictures; The program automatically downloads and configures the model, without the need for GPU configuration and network dependency, and can be generated as many times as possible, and can be licensed for free for commercial use. Built-in multi-style rendering and parameter adjustment such as negative prompt words, resolution, and sampling steps to meet the design needs of illustrations, posters, e-commerce, social media, etc.; The local operation ensures data privacy, is lightweight, and can be used immediately after decompression, greatly reducing the threshold for AI mapping, and helping individuals and teams efficiently produce original visual content.

Turbo.Art is an AI drawing platform from Modal Labs that provides real-time intelligent image generation and editing capabilities based on SDXL Turbo models based on Stability. Users can freely paint on the blank canvas or enter text prompts to generate multi-style high-definition paintings with one click, supporting a variety of art styles such as cyberpunk, realistic, cartoon and Miyazaki. The platform's built-in enhancement algorithms allow for local fine-tuning of generated results and provide high-quality PNG/JPEG exports. The front-end code is open source on GitHub, supports API calls and custom deployments, and helps designers, content creators, and developers efficiently produce original visual content in advertising, social media, and art.

Stable Diffusion WebUI is a browser-based AI image generation platform that allows users to quickly create high-quality images with natural language prompt words. The platform integrates a variety of functions, including text-to-image (txt2img), image-to-image generation (img2img), image inpainting, and image extension (outpainting), etc., to meet the diverse needs of users in art creation, design sketches, advertising creatives, etc. Users do not need local deployment or programming knowledge, and can realize AI drawing by simply operating through the web interface. The platform supports a variety of model and plug-in extensions, and provides flexible parameter settings, which is suitable for designers, content creators, and AI enthusiasts to improve creative efficiency and visual expression.