AI PLAYGROUND

Welcome to AI Playground open source project and AI PC generative AI application suite. This application provides a full suite of generative AI features for chat, code assistance, document search, image analysis, image and video generation. All features run offline and are powered by your PC’s Intel® Core™ Ultra with built-in Intel Arc GPU or Intel Arc™ dGPU Series A or B with 8GB+ of vRAM.

AI Playground is intended to act as an offline alternative to cloud tools such Gemini ChatGPT and Grok. AI Playground leverages libraries from GitHub and Huggingface including:

• Image Diffusion (PyTorch): Stable Diffusion 1.5, SDXL, Flux.1-Schnell, Flux.1 Kontext[dev], Z-Image, Wan2.1 VACE, LTX-Video
• LLM: GGUF (Llama.cpp Vulkan) - Gemma4, Qwen3.5, Qwen3 VL, GPT-OSS 20B, DeepSeek R1 Distilled, Phi3, Mistral 7B, Llama 3.2: OpenVINO - TinyLlama, Mistral 7B, Phi3 mini, Phi3.5 mini, DeepSeek R1 Distill (1.5B, 7B)

As a local alternative to cloud AI service, AI Playground is intended to give consumers and AI curious prosumers 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 3rd 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 Gemma4, Qwen3.5, 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 Analyze images with Qwen3 VL Model Vibe Coding with GPT-OSS 20B Reasoning Document RAG with Mistral 7B Instruct
• 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.

README.md

• English (readme.md)

Min Specs

AI Playground alpha and beta installers are currently available downloadable executables, or available as a source code from our Github repository. To run AI Playground you must have a PC that meets the following specifications

• Windows OS
• Intel Core Ultra Series 3, Series 2H, Series 2V, or Series 1 H processor OR Intel Arc GPU Series A or Series B (discrete) with 8GB of vRAM, or Nvidia RTX GPU

Installation - Packaged Installer:

This is a single packaged installer for all supported hardware mentioned above. This installer simplifies the process for end users to install AI Playground on their PCs. Please note that while this makes the installation process easier, this is open-source beta software, and there may be component and version conflicts. Refer to the Troubleshooting section for known issues.

Download the installer

🆕 AI Playground 3.1.0 alpha_hf (all SKUs) - Release Notes | Download 🆕

Installation Process for v3.x

1. The installer only installs the Electron frontend, so it completes very quickly.
2. On the first run, AI Playground Setup window appears where you can select from hardware modes and needed backend components for AI Playground to function properly. This process requires a strong and open network and may take several minutes.
3. Download the Users Guide for application information: AI Playground Users Guide

Troubleshooting Installation

The following are known situations where your installation may be blocked or interrupted. Review the following to remedy installations issues. If installation issues persist, generate a copy of the log by typing CTRL+SHIFT+I, select the console tab and copy the last few entries of the log written where the installer failed. Provide these details to us via the issues tab here, or via the Intel Insiders Discord, or Graphics forum on Intel's support site.

1. Llama.cpp embedding issues: At the time of this release, Llama.cpp embeddings may have issues with:

• Recent drivers, and may require DDU to clean driver cache.
• Anti-Virus software - features needed to read and write embedding cache may not be properly installed: Disable anti-virus, restart

2. Restart: Time-out issues have been sighted, which show as a failed install but resolve when restarting AI Playground
3. Verify Intel Arc GPU: Ensure your system has an Intel Arc GPU with the lastest driver. Go to your Windows Start Menu, type "Device Manager," and under Display Adapters, check the name of your GPU device. It should describe an Intel Arc GPU. If so, then you you have a GPU that means our minimum specifications. If it says "Intel(R) Graphics," your system does not have a built-in Intel Arc GPU and does not meet the minimum specifications. If your GPU is an discrete GPU such as Intel Arc A or B series GPU, then you can troubleshoot a troubled installation by disabling the iGPU in Device Manager
4. Interrupted Installation: The online installation for backend components can be interrupted or blocked by an IT network, firewall, or sleep settings. Ensure you are on an open network, with the firewall off, and set sleep settings to stay awake when powered on.
5. Missing Libraries: Some Windows systems may be missing needed libraries. This can be fixed by installing the 64-bit VC++ redistribution from Microsoft here. It is recommended this be done after updating the Graphics drivers. Then install AI Playground.
6. Python Conflict: Some PCs with an existing installation of Python can cause a conflict with AI Playground installation, where the wrong or conflicting packages are installed due to the incorrect version or location of Python on the system. This is usually remedied by uninstalling Python environment, restarting and reinstalling AI Playground
7. Temp Files: Should the installation be interrupted because of any of the above issues it is possible that temporary installation files have been left behind and trying to install with these files in place can block the installation. Remove these files or do a clean install of AI Playground to remedy

Project Development

Checkout Source Code

To get started, clone the repository and navigate to the project directory:

git clone -b dev https://github.com/intel/AI-Playground.git
cd AI-Playground

Install Node.js Dependencies

1. Install the Node.js ≥ 20 development environment.

   • Windows / macOS: download the installer from Node.js.

   • Ubuntu 24.04 / Debian: the distro's apt install nodejs ships Node 18 (no npm) and is too old. Install Node 22 LTS via NodeSource instead:

     curl -fsSL https://deb.nodesource.com/setup_22.x | sudo -E bash -
     sudo apt install -y nodejs
     node --version    # expect: v22.x
     npm  --version    # expect: 10.x

     Behind a corporate proxy that intercepts deb.nodesource.com, prefix the curl with --proxy "$http_proxy" and ensure https_proxy is exported.

2. Navigate to the WebUI directory and install all Node.js dependencies:

cd WebUI
npm install

Fetch External Resources

1. In the WebUI directory, execute the fetch-external-resources script to download required external resources:

This will download uv (Python package manager) and other required tools to the build/resources/ directory.

Launch the application

To start the application in development mode, run:

npm run dev

(Optional) Build the installer

To build the installer, run:

npm run build

The installer executable will be located in the build/electron folder.

Build and run on Linux (AppImage or .deb)

Linux support is experimental. The frontend, AI Backend, LlamaCPP and ComfyUI backends run on Ubuntu x64. The packaged installer/AppImage/.deb supports Ubuntu 24 or newer only. See docs/linux-intel-gpu-setup.md for GPU driver requirements.

The Linux build produces both a single, portable AppImage (no installation, no root) and a .deb package for a system-wide install via apt.

1. Build it from the WebUI directory:

   cd WebUI
   npm install                        # install build dependencies
   npm run fetch-external-resources   # one-time: downloads uv/7zip
   npm run build:linux

   Both AI Playground-<version>.AppImage and AI Playground-<version>.deb are written to build/electron. The build host needs ar (from binutils, usually already present) to assemble the .deb.

2. Install FUSE once (required to run any AppImage):

   sudo apt install -y libfuse2t64

3. Make it executable and start the app:

   cd build/electron
   chmod +x "AI Playground-"*.AppImage
   ./"AI Playground-"*.AppImage

   --no-sandbox is baked into the launcher (see build/scripts/after-pack.cjs), so you can also start the app by double-clicking the AppImage in your file manager — provided the file manager is allowed to run executables (in GNOME Files: Preferences → "Executable Text Files" / right-click → Run, or chmod +x as above). Do not start it with sudo — Electron refuses to run as root.

   Alternatively, install the .deb instead of running the AppImage. apt resolves the runtime dependencies declared in the package (libgtk-3-0, libnss3, libasound2, libdbus-1-3, pciutils, python3):

   cd build/electron
   sudo apt install ./"AI Playground-"*.deb

   This installs the app system-wide; launch it as AI Playground from your application menu or run ai-playground from a terminal. --no-sandbox is baked into the launcher here too. Uninstall with sudo apt remove ai-playground.

4. OpenVINO Ubuntu dependencies are checked during OpenVINO backend setup.

   If required packages are missing, AI Playground can open a terminal installer. The terminal asks for your sudo password and runs the package install command.

   On Ubuntu, the install command is:

   sudo apt-get update
   sudo apt-get install -y python3 python3-venv libtbb12 libhwloc15 libgomp1 libnuma1 ocl-icd-libopencl1 libfuse2t64

Behind a corporate / HTTP proxy

AI Playground downloads tools at build time (uv, 7zip) and backend binaries on first run (llama.cpp, OpenVINO/OVMS). Both honor the standard proxy environment variables when they are set — export them before building or launching:

export https_proxy="http://proxy.example.com:port"
export http_proxy="http://proxy.example.com:port"
export no_proxy="localhost,127.0.0.1"   # hosts to reach directly

With these set:

• Build (npm run fetch-external-resources / npm run build:linux) routes Node's downloads through the proxy (the script runs with NODE_USE_ENV_PROXY=1).
• Runtime the app reads the same variables at startup and points Electron's network stack (net.fetch) at the proxy, so llama.cpp and OVMS downloads succeed.

⚠️ Launching from a file manager won't pick up the proxy. Double-clicking the AppImage starts it from the desktop session, which does not inherit http_proxy exported in ~/.profile or ~/.bashrc. Either launch the AppImage from a terminal where the variables are exported, or configure a system-wide proxy (e.g. GNOME Settings → Network → Network Proxy) so the desktop session provides them.

If your network blocks the downloads entirely, you can also pre-place the build tools manually: drop the extracted uv/7zip binaries into WebUI/build/resources/ (the fetch script skips any file already present) before running the build.

Corporate proxy + Intel-owned external hosts

A common gotcha on Intel networks: apt and curl honor no_proxy exactly, but a broad entry like no_proxy=*.intel.com will also match the externally-hosted CDN repositories.intel.com (resolves to AWS), causing apt update to hang trying to reach it directly through the corporate firewall.

Work around it by scoping the proxy override per-host rather than rewriting no_proxy:

# /etc/apt/apt.conf.d/99-intel-proxy
Acquire::http::Proxy::repositories.intel.com  "http://proxy.example.com:port";
Acquire::https::Proxy::repositories.intel.com "http://proxy.example.com:port";

For interactive curl calls (e.g. fetching the NPU driver tarball from github.com which redirects to AWS), pass --proxy "$http_proxy" explicitly.

Model Support

AI Playground does not ship with any generative AI models but does make models available for all features either directly from the interface or indirectly by the users downloading models from HuggingFace.co or CivitAI.com and placing them in the appropriate model folder.

Models currently linked from the application

AI Model & License Registry

Model Path / Name	Model Card (HF)	License Link
AdamCodd/vit-base-nsfw-detector	Model Card	Apache 2.0
Aitrepreneur/insightface/inswapper_128.onnx	Model Card	Non-Commercial
alimama-creative/FLUX.1-Turbo-Alpha	Model Card	FLUX.1-dev License
BGE Small EN v1.5 (GGUF)	Model Card	MIT License
black-forest-labs/FLUX.2-klein-4b-fp8/flux-2-klein-4b-fp8.safetensors	Model Card	Apache 2.0
city96/t5-v1_1-xxl-encoder-gguf/t5-v1_1-xxl-encoder-Q3_K_M.gguf	Model Card	Apache 2.0
city96/t5-v1_1-xxl-encoder-gguf/t5-v1_1-xxl-encoder-Q4_K_M.gguf	Model Card	Apache 2.0
city96/umt5-xxl-encoder-gguf/umt5-xxl-encoder-Q4_K_M.gguf	Model Card	Apache 2.0
comfyanonymous/flux_text_encoders/clip_l.safetensors	Model Card	Apache 2.0
comfyanonymous/flux_text_encoders/t5xxl_fp8_e4m3fn_scaled.safetensors	Model Card	Apache 2.0
Comfy-Org/flux1-kontext-dev/flux1-dev-kontext_fp8_scaled.safetensors	Model Card	FLUX.1-dev License
Comfy-Org/Lumina_Image_2.0_Repackaged/ae.safetensors	Model Card	Apache 2.0
Comfy-Org/Real-ESRGAN_repackaged/RealESRGAN_x4plus.safetensors	Model Card	BSD-3-Clause
Comfy-Org/Wan_2.1_ComfyUI_repackaged/wan_2.1_vae.safetensors	Model Card	Wan 2.1 License
Comfy-Org/z_image_turbo/ae.safetensors	Model Card	Apache 2.0
Comfy-Org/z_image_turbo/qwen_3_4b.safetensors	Model Card	Apache 2.0
Comfy-Org/z_image_turbo/z_image_turbo_bf16.safetensors	Model Card	Apache 2.0
DeepSeek-R1-Distill-Qwen 1.5B	Model Card	MIT License
DeepSeek-R1-Distill-Qwen 7B	Model Card	MIT License
Gemma 3 4B IT (Unsloth)	Model Card	Gemma License
gmk123/GFPGAN/GFPGANv1.4.pth	Model Card	Apache 2.0
GPT-OSS 20B (Unsloth)	Model Card	Apache 2.0
InternVL2 4B (OV)	Model Card	Apache 2.0
latent-consistency/lcm-lora-sdv1-5/pytorch_lora_weights.safetensors	Model Card	OpenRAIL++
latent-consistency/lcm-lora-sdxl/pytorch_lora_weights.safetensors	Model Card	OpenRAIL++
Lightricks/LTX-Video/ltxv-2b-0.9.6-distilled-04-25.safetensors	Model Card	Apache 2.0
Llama 3.2 3B Instruct	Model Card	Llama 3.2 License
lllyasviel/fooocus_inpaint/fooocus_inpaint_head.pth	Model Card	OpenRAIL
lllyasviel/fooocus_inpaint/inpaint_v26.fooocus.patch	Model Card	OpenRAIL
Lykon/DreamShaper/DreamShaper_8_pruned.safetensors	Model Card	OpenRAIL-M
Lykon/dreamshaper-8-inpainting/text_encoder/model.safetensors	Model Card	OpenRAIL-M
Lykon/dreamshaper-8-inpainting/unet/model.safetensors	Model Card	OpenRAIL-M
Lykon/dreamshaper-8-inpainting/vae/model.safetensors	Model Card	OpenRAIL-M
Meta-Llama 3.1 8B Instruct	Model Card	Llama 3.1 License
Mistral 7B Instruct v0.2 (OV)	Model Card	Apache 2.0
Mistral 7B Instruct v0.3	Model Card	Apache 2.0
Mistral 7B Instruct v0.3 (OV)	Model Card	Apache 2.0
Nomic Embed Text v1.5 (GGUF)	Model Card	Apache 2.0
Phi-3 Mini 4k Instruct (OV)	Model Card	MIT License
Phi-3.5 Mini Instruct (OV)	Model Card	MIT License
QuantStack/Wan2.1_14B_VACE-GGUF/Wan2.1_14B_VACE-Q8_0.gguf	Model Card	Apache 2.0
Qwen2-VL 7B Instruct (OV)	Model Card	Apache 2.0
Qwen3 4B (OV)	Model Card	Apache 2.0
Qwen3 4B (Unsloth)	Model Card	Apache 2.0
Qwen3 4B Instruct 2507 (Unsloth)	Model Card	Apache 2.0
Qwen3-VL 4B Instruct (Unsloth)	Model Card	Apache 2.0
RunDiffusion/Juggernaut-XL-v9/RunDiffusionPhoto_v2.safetensors	Model Card	OpenRAIL-M
SmolLM2 1.7B Instruct	Model Card	Apache 2.0
stabilityai/control-lora/rank128-canny-rank128.safetensors	Model Card	SAI Community
tencent/Hunyuan3D-2.1/hunyuan3d-dit-v2-1/model.fp16.ckpt	Model Card	Hunyuan3D License
tencent/Hunyuan3D-2/hunyuan3d-dit-v2-0/model.fp16.safetensors	Model Card	Hunyuan3D License
TinyLlama 1.1B Chat (OV)	Model Card	Apache 2.0
Whisper (OV)	Model Card	Apache 2.0

Be sure to check license terms for any model used in AI Playground especially taking note of any restrictions.

Use Alternative Models

Check the User Guide for details or watch this video on how to add alternative Stable Diffusion models to AI Playground

Notices and Disclaimers:

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

The software may include third party components with separate legal notices or governed by other agreements, as may be described in the Third Party Notices file accompanying the software.

Credit

License details for borrowed code and components can be found in our 3rdpartynoticeslicense file.
Additionally, these entities and their work stand out as are fundamental to AI Playground.

• PyTorch - https://pytorch.org/
• Stable Diffusion - https://github.com/Stability-AI/stablediffusion
• ComfyUI - https://github.com/comfyanonymous/ComfyUI
• OpenVINO - https://openvinotoolkit.github.io/openvino.genai/
• Llama.cpp - https://github.com/ggml-org/llama.cpp
• Vue.js - https://vuejs.org/
• Plus countless other open-source projects and contributors that make this work possible!