Introducing AI Playground

Updated 01/07/2026

AI Playground 是一套易於使用的 Generative AI 應用程式套件，適用於 Intel AI PC，採用 Intel Arc GPU，可內建 Intel® Core™ Ultra 處理器或透過 Intel® Arc™ 獨立 GPU。 AI Playground 可以開放原始碼專案及 Windows 桌面安裝套件的形式提供。

全新: An early alpha preview of AI Playground 3.0 is now available with support for vision models like Qwen3 VL, agentic tool calling for generating across features in a single chat discussion, and speech to prompt mode using the Whisper model and much more.  Still fresh in AI Playground 2.6.2. is support for GPT-OSS via Llama.cpp Vulkan, PyTorch 2.8 with support for Flux.1 Kontext[dev], and Wan 2.1 VACE video gen, plus OpenVINO for chat (for GPU or NPU), and continued support for DeepSeek models, SD1.5 SDXL, Flux.1-Schnell and ComfyUI workflows, and more.

Download AI Playground: 

Be sure to update your Graphics and NPU drivers(for supported Intel Core Ultra devices).  Also note while an installer makes the process easy to get up and running, this is open source beta software, where components and version can have conflicts. Check the 疑難排解部分 的專案 readme 中的已知問題。

安裝前，先備份先前安裝時所產生的任何有價值內容

For information and features on early access AI Playground v.3.0.0 alpha see the branch source readme and manual. This early alpha preview is rebuilt from the ground up with new agentic AI features, vision model support, and updated image generation and editing capabilities.  A beta release is expected in Q1 2026

 

What is AI Playground

AI Playground beta is a free to download and app for doing Gen AI inferencing on your Intel Core Ultra or Intel Arc GPU based systems. AI Playground curates over 60 top open-source models and ties them to a variety of point and click features for chat, image generation, image generation and video.

Easy To Use:AI Playground 的設計簡單易用，讓您可以快速使用 AI。AI Playground 會在任務需要模型時提示您，讓您下載或提供自己的模型。而且我們的設定是使用對任務合理的術語來撰寫和設計的，您不需要為了使用 AI 而接受 AI 教育。

All-In-One AI Hub: 無縫切換 影像建立, 影像編輯和一個 聊天 LLM .AI Playground 將這些功能整合在一個易於使用的體驗中。目前 AI Playground 支援 Stable Diffusion 1.5、Stable Diffusion XL、Flux.1-Schnell、LCM Fast LoRAs、Image Upscale、Image to Image Stylization、Inpainting、Outpainting、ChatBot with session history context 和 LLM RAG 功能，讓您可以從本機電腦取得摘要和搜尋內容。

Flexible: AI Playground will provide access to models for all features, and more advanced users can install their favorite models and tweak manual settings, to configure AI Playground to work the way they want it to work.

Local:AI Playground 在本機執行，因此您的提示和輸出從未離開您的電腦。這提供您隱私權，並控制您與 AI 合作的內容與輸入，讓您能掌控全局，而不會有上游實體在您的肩上審查您的提示或擷取您的資料與內容。

Hardware Optimized:AI Playground 採用 Intel Arc GPU 架構，利用我們強大的 Intel Xe 核心和 XMX 引擎。

Open: AI Playground is also an open-source project where our community can clone, fork or help us address features and capabilities of the main repo.

核心功能

AI Playground is a locally run Generative AI application suite that acts as an offline alternative to cloud tools such Gemini ChatGPT and Grok.  AI Playground is intended to give consumers and AI curious prosumer easy and intuitive access to a wide variety of generative AI features using their Intel powered AI PC. This means you can be offline, without loading sensitive or personal data to 3^rd party sites, for free, in a single app without having to know how to install and manage multiple AI backend frameworks.   Key features:

• Latest and greatest chat models: Support for Qwen 3 VL, Mistral 7B, DeepSeek R1 or GPT-OSS, AI playground makes a variety of chat models available to users
• Vision, Reasoning and RAG: Chat features support Vision, Reasoning and RAG to analyze and get deep answers on both visual and text content.
  
• Image Generation: From Stable Diffusion 1.5, SDXL, Flux.1 and Z-image models AI Playground is making a breadth of image generation from quick easy low-res draft generation to high quality image generation
• Image Editing: Subscription free and private control for upscaling, inpainting, outpainting, 2D to 3D mesh or editing images in a variety of ways. Good for editing personal photos to taking sketches and generated images to the next level with greater control.
   

New To AI Playground

AI Playground 提供一系列新的解決方案，每項解決方案都能擴充 AI Playground 的功能，以其模組化架構為基礎，即使在不同版本之間也能提供價值與功能。

New Modalities

• Voice mode: AI Playground now supports a voice mode. From the app settings you can turn on Voice mode. This will load an AI voice model to the NPU of your Intel Core Ultra PC. When enabled a microphone icon will be available in the prompt. Simple click the icon and voice your prompt. Voice mode will convert your voice to text prompt
• Vision Model Support: New to AI Playground is Vision Model support. With Vision support you can paste or load an image to Chat mode and have AI Playground analyze the image, giving you insight, answers and information about the image
• Agentic Multi-Modal Tools. AI Playground exposes MCP tooling to our chat mode allowing you to use a core model to do agentic tasks across AI Playground features within a single chat discussion. This feature will allow you to use text chat, vision chat, and image generation withing a single discussion window.  See Agentic Preset for more information

Unified Prompt:

AI Playground now deploys a single unified prompt window for all AI Playground features. This design aligns AI Playground to familiar AI tools like Gemini, Chat GPT and Grok

Light Theme:

You can now choose to experience AI Playground in a light-colored theme

Prompt Modes, Settings and Presets:

The main prompt can now be directed to specific models and tasks through the Prompt Mode, Prompt Settings and Prompt Presets.

• Prompt Modes allow you to quickly and easily change between modalities from the main prompt: Select chat for vision and text based chat responses, Image Gen for image generation from text. Image Edit for altering or enhancing images. Video Gen: For creating video content from text or image input
• Prompt Mode Presets: From the Prompt Settings button you have access to dozens of Presets. Presets are similar to the previous AI Playground “Workflows” Settings, configured for specific tasks and usages but are now expanded to all Prompt modes. With Presets you can quickly access a define usage for Chat, Image Gen, Image Edit or Video where the model selections and settings are preset for that task. See the Preset Section for more information

Backends:

• OpenVINO: OpenVINO provides the best performance when using OpenVINO IR Models with the ability to target CPU, NPU or GPU. AI Playground provides access to a variety of chat, reasoning, vision and tool calling models in OpenVINO format in Chat mode:
• Llama.CPP:  For access to latest models in GGUF format, AI Playground provides models such as Qwen 3VL, DeepSeek R1 Distilled, and GPT-OSS 20B in GGUF format, for use in the chat mode
• PyTorch: PyTorch is primarily used to support image diffusion based models for image generation, image editing and video generation. We provide access to models such as SDXL, Flux.1 Schnell, Z-Image-Turbo and more
• ComfyUI: is a backend solution allowing us to easily create and deploy custom workflows for specific image and video Preset 

ComfyUI 編輯器

For users interested in diving into ComfyUI workflows, or those already familiar with ComfyUI, we’ve provided direct access to its workflow editor, allowing users to edit or create workflows for use either directly from ComfyUI or in AI Playground. We encourage and welcome the community created workflows submitted and shared to AI Playground via Pull Requests to our project

See AI Playground In Action

AI Playground 3.0 at CES 2026

示範錄影帶

Introduction Video of Core Features

Install Image Checkpoints

For more information, visit the AI Playground project on GitHub, or join our Intel Insiders Discord to engage directly with Intel and our Community on AI Playground

Notices and Disclaimers: For information on AI Playground terms, license and disclaimers, visit the project and files on GitHub repo:

License | 備註與免責聲明