Modern AI models and tools are transforming real-world GenAI applications, enabling powerful capabilities for image generation, stylizing, and chatbot interactions. What if you could explore and experiment with the latest GenAI models and tools in a fully open, user-friendly, secure, and no-cost environment without requiring a network connection? AI Playground is the solution that makes it possible! It is a comprehensive suite of GenAI apps available as an open-source project on GitHub. The latest version supports Intel® Core™ Ultra 200H Processors and PyTorch* 2.7. It comes with an enhanced chat feature backed by the OpenVINO™ toolkit and text-to-video conversion powered by the Lightricks* LTX-Video model. Furthermore, it facilitates AI developers with Lightricks ComfyUI workflows, DeepSeek models on Hugging Face*, Hugging Face GGUF format for LLMs, and much more!
|
In the video ‘AI Playground: Text Generation, RAG and more’ available on YouTube, Guy Tamir briefly introduces AI Playground and shows how it enables GenAI applications like text generation and RAG while using the models of your choice. Watch the video below for more details.
Why AI Playground?
AI Playground is an all-in-one hub for AI developers that integrates various GenAI applications, including image generation, image editing, and AI chat functionalities, using Retrieval Augmented Generation (RAG) into a single AI PC starter app. It supports many popular GenAI libraries and models, such as Meta* Phi3, Meta Llama, Mistral 7B, Qwen2, and more. It provides developers with the flexibility to choose between simple and advanced AI models depending on their needs and expertise. They also have the flexibility to adjust model settings, enabling a high level of customization.
Its support for advanced image generation models like Stable Diffusion XL and Flux.1-Schnell allows the creation of detailed and realistic images from text prompts, with options for fine-tuning and customization using powerful image editing tools. Advanced features like in-painting and out-painting enable users to make sophisticated adjustments to their creations.
In the demo video ‘AI Playground: Image Generation and Enhancement’ on YouTube, Guy demonstrates a few examples on how you can generate the images of your choice on preferred hardware (see the release notes for a list of supported architectures), and further enhance them by providing appropriate text prompts using the image editing features of the app. More details are in the video below.
The AI chatbot feature in AI Playground includes session history context and LLM RAG capabilities, which provides users which the ability to summarize and search local content. This makes the chatbot more effective and responsive to user queries.
AI Playground is designed to be user-friendly, with intuitive interfaces and workflows that guide users through the process of generating and editing images or interacting with AI chatbots. This makes the app accessible to both beginners and advanced users. One of the AI Playground’s standout features is that all operations are performed locally on your PC. This ensures that prompts and outputs never leave your computer, enhancing data privacy and security.
The open-source availability of the AI Playground project encourages transparency and community contributions. Users can download a packaged Windows desktop installer and customize the application to suit their preferences.
|
Watch the video ‘AI Playground is Open Source’ on YouTube.
Intel strongly encourages open-source software development by actively contributing to the open AI framework and developer communities, such as Python*, PyTorch, TensorFlow*, OpenVINO, and more. We also upstream our optimizations for developers to securely leverage the open standards across the latest heterogeneous architectures for advanced AI and accelerated computing. Learn more about our commitment to open-source code development here.
AI Playground has been built reflecting our desire to enable responsible use of AI through an open, innovative ecosystem. It leverages the efforts we actively put into open AI framework communities such as PyTorch and OpenVINO for accelerated and optimized GenAI.
Enhanced GenAI Backed by PyTorch* and OpenVINO™
The PyTorch AI framework forms the backbone of the AI Playground project. Developers can seamlessly integrate existing features and optimizations of PyTorch into their workflows on AI Playground, and thus focus more on innovation rather than worrying about the underlying infrastructure. AI Playground leverages eager mode of PyTorch, enabling faster prototyping and iterations on advanced AI features. On top of that, the integration of Intel® Extension for PyTorch into the platform helps increase the efficiency and scalability of GenAI workloads on Intel® hardware.
Recent versions of AI Playground come with OpenVINO as a chat backend option. This preview feature enables fast and efficient inference on Intel hardware.
Go Play
Get started with AI Playground, a robust and versatile platform for exploring GenAI. Its range of features caters to both novice and experienced users. Its ease of use, local operation, high-quality outputs, and community support make it a valuable platform for anyone interested in AI-driven content creation.
Useful Resources
- AI Playground: GitHub Project
- AI with Guy: YouTube Playlist
- Introduction to AI Playground: YouTube Video
- Running AI on Intel Arc GPUs: YouTube Video
- AI Playground Users Guide
- PyTorch Optimizations from Intel
- Jump-start AI Development
- GenAI Development Resources
Technical Software Product Marketing Engineer, Intel
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.