Releases: intel/AI-Playground
Releases · intel/AI-Playground
Release list
AI Playground Version v3.1.2 - hot fix 3
AI Playground Version v3.1.2 hot fix 3
This release is a hot fix to AI Playground 3.1.2 beta.
Updates To This Release
- PDF document fix
- Version 3.1.2 beta hf2 updates See release
- Version 3.1.2 beta hf updates - See release
- Version 3.1.2 beta updates - See release
Version 3.1.2 beta Supported Hardware
- Intel Core Series 3 Processors with 12GB+ of system memory (WCL)
- Intel Core Ultra Series 3 Processors (PTL)
- Intel Core Ultra Series 2 (H) Processors (ARL-H)
- Intel Core Ultra Series 2 (V) Processors (LNL)
- Intel Core Ultra Series 1 (H) Processors (MTL-H)
- Intel Arc B Series GPU Cards (BMG)
- Intel Arc A Series GPU Cards with 8GB+ Memory (ACM)
- Nvidia RTX GeForce GPUs
- Home Agent feature requires a minimum of 32GB RAM laptops, or 16GB on discete GPU
AI Playground v3.1.2 beta - hot fix2
AI Playground Version v3.1.2 hot fix 2
This release is a hot fix to AI Playground 3.1.2 beta.
Updates To This Release
- PyTorch version contained to 2.12.x for broader compatibility
- Version 3.1.2 beta hf updates - See release
- Version 3.1.2 beta updates - See release
Version 3.1.2 beta Supported Hardware
- Intel Core Series 3 Processors with 12GB+ of system memory (WCL)
- Intel Core Ultra Series 3 Processors (PTL)
- Intel Core Ultra Series 2 (H) Processors (ARL-H)
- Intel Core Ultra Series 2 (V) Processors (LNL)
- Intel Core Ultra Series 1 (H) Processors (MTL-H)
- Intel Arc B Series GPU Cards (BMG)
- Intel Arc A Series GPU Cards with 8GB+ Memory (ACM)
- Nvidia RTX GeForce GPUs
- Home Agent feature requires a minimum of 32GB RAM laptops, or 16GB on discete GPU
AI Playground v3.1.2 beta - hot fix
AI Playground Version v3.1.2 hot fix
This release is a hot fix to AI Playground 3.1.2 beta.
Updates To This Release
- Remove hard coded ComfyUI parameter on Linux unnecessary
- Probe XPU before Linux spawn instead of forcing CPU
- Inswapper bug fix
AI Playground v3.1.2 beta
AI Playground Version v3.1.2 beta
AI Playground 3.1.2 beta is an incremental update to AI Playground 3.1.1 alpha, with adjusted features, fixes and moderate feature additions from the previous release.
Version 3.1.2 beta Updates
- Ubuntu Linux .deb package (video gen currently unsupported)
- Home Agent: This feature now un-hidden as a setup option (32GB+ recommended)
- Home Agent: Slack added as a Home Agent messaging app
- Home Agent: Change back-end, change model, change context and change token skills added to Home Agent
- Video and 3D Gen features added to agentic tool calling, also available from Home Agent
- Refactored built-in tools, and added new web search and screenshot tools for Agentic Chat and Home Agent features
- Phison aiDAPTIV software updated, and chat preset added supporting a list of large MOE models to install and run
- Text To Speech feature added to app settings with speech output playable from chat results, and speech output files automatically provided to mobile Home Agent chats when speech is recorded for prompts
- Qwen 3.5 4B and Qwen 3.5 9B models added for OpenVINO
Known Issues
- Linux version crashes with video generation. Working on fix with upcoming hot fix release
- Vulkan instability seen after demanding tool calling, reverting to OpenVINO, or clearing shader cache can resolve
- The installer can fail due to environmental issues. See the Troubleshooting guide for help
- Systems with discrete GPU and integrate GPU may need to disable iGPU when installing
- Systems with more than one GPU may need to disable idle GPUs when inferencing
- An interrupted download can block models from downloading properly. To fix this, go to the model location, delete the temporary file, and try again.
- The Colorizer workflow is limited to Xe2-supported GPUs (LNL and BMG)
- A PC restart may be needed after installation for optimal functionality and performance
- Reinstalling over an existing AI Playground installation will delete any ComfyUI nodes.
- Users can adjust settings beyond their specific system's ability resulting in failed generations
Version 3.1.2 beta Supported Hardware
- Intel Core Series 3 Processors with 12GB+ of system memory (WCL)
- Intel Core Ultra Series 3 Processors (PTL)
- Intel Core Ultra Series 2 (H) Processors (ARL-H)
- Intel Core Ultra Series 2 (V) Processors (LNL)
- Intel Core Ultra Series 1 (H) Processors (MTL-H)
- Intel Arc B Series GPU Cards (BMG)
- Intel Arc A Series GPU Cards with 8GB+ Memory (ACM)
- Nvidia RTX GeForce GPUs
- Home Agent feature requires a minimum of 32GB RAM laptops, or 16GB on discete GPU
AI Playground v3.1.1 beta
AI Playground Version v3.1.1 beta
AI Playground 3.1.1 beta is an incremental update to AI Playground 3.1.0 alpha, with adjusted features, fixes and moderate feature additions from the previous release.
Version 3.1.1 beta Updates
- New Pro 3 Image Preset using ERNIE Image models
- New Video Gen Presets for LTX 2.3 t2v/i2V, and WAN 2.2 i2v
- Added OpenVINO support for AI Playground Studio using GPU, for Image Gen (Draft, HD) and Image Edit (Upscale, Inpaint, Outpaint) presets
- Acer MCP integration for Agentic Chat preset (installs on detected Acer systems)
- Phison Pascari aiDAPTIV™ Cache Memory SSD support for run large MOE models (ie GPT-OSS 120B on 32GB Intel Core Ultra PC)
- Experimental early release of remote AI Playground Home Agent using Telegram chat mobile app
(enable through settings.json)
Known Issues
- The installer can fail due to environmental issues. See the Troubleshooting guide for help
- Systems with discrete GPU and integrate GPU may need to disable iGPU when installing
- Systems with more than one GPU may need to disable idle GPUs when inferencing
- An interrupted download can block models from downloading properly. To fix this, go to the model location, delete the temporary file, and try again.
- The Colorizer workflow is limited to Xe2-supported GPUs (LNL and BMG)
- A PC restart may be needed after installation for optimal functionality and performance
- Reinstalling over an existing AI Playground installation will delete any ComfyUI nodes.
- Users can adjust settings beyond their specific system's ability resulting in failed generations
Version 3.1.1 beta Supported Hardware
- Intel Core Series 3 Processors with 12GB+ of system memory (WCL)
- Intel Core Ultra Series 3 Processors (PTL)
- Intel Core Ultra Series 2 (H) Processors (ARL-H)
- Intel Core Ultra Series 2 (V) Processors (LNL)
- Intel Core Ultra Series 1 (H) Processors (MTL-H)
- Intel Arc B Series GPU Cards (BMG)
- Intel Arc A Series GPU Cards with 8GB+ Memory (ACM)
- Nvidia RTX GeForce GPUs
AI Playground v3.1.0 alpha - hot fix
AI Playground Version v3.1.0 alpha hot fix
Important
This is a hot fix release to AI Playground 3.1.0 alpha to address a fix ComfyUI installation error related to the latest PyTorch release. The rest of this release matches 3.1.0 alpha with the following release information.
AI Playground 3.1.0 alpha an early alpha release and may have minor bugs or breaks in the user experience, as it is adding new version options not previously available in AI Playground.
Note: This release's installer binary is larger than usual, containing a needed oneDNN update, scheduled to be release in PyTorch 2.12. A smaller installer binary will be released once that distribution is through PyTorch.
Version 3.1.0 alpha Updates
- Hardware Mode installer - detects hardware and provides targeted installation
- AI Playground Essentials - a curated installation of features for Intel Core Series 3 Processors
- AI Playground Studio - a named version for the standard AI Playground features targeting Intel Core Ultra Processors and Intel Arc GPUs
- AI Playground Studio, NVIDIA CUDA - Experimental installation option to run local inferencing on Nvidia GPUs via torch.cuda
- Agentic MCP Support - (for studio versions), use AI Playground Agentic chat as host MCP for 3rd party MCP Server applications
- Alternative Demo mode and features for Essentials
- Gemma 4 E4B added to Llama.cpp chat
Known Issues
- The installer can fail due to environmental issues. See the Troubleshooting guide for help
- Systems with discrete GPU and integrate GPU may need to disable iGPU when installing
- Systems with more than one GPU may need to disable idle GPUs when inferencing
- An interrupted download can block models from downloading properly. To fix this, go to the model location, delete the temporary file, and try again.
- The Colorizer workflow is limited to Xe2-supported GPUs (LNL and BMG)
- A PC restart may be needed after installation for optimal functionality and performance
- Reinstalling over an existing AI Playground installation will delete any ComfyUI nodes.
- Users can adjust settings beyond their specific system's ability resulting in failed generations
Version 3.1.0 alpha Supported Hardware
- Intel Core Series 3 Processors with 12GB+ of system memory (WCL)
- Intel Core Ultra Series 3 Processors (PTL)
- Intel Core Ultra Series 2 (H) Processors (ARL-H)
- Intel Core Ultra Series 2 (V) Processors (LNL)
- Intel Core Ultra Series 1 (H) Processors (MTL-H)
- Intel Arc B Series GPU Cards (BMG)
- Intel Arc A Series GPU Cards with 8GB+ Memory (ACM)
- Nvidia RTX GeForce GPUs
AI Playground v3.1.0 alpha
AI Playground Version v3.1.0 alpha
Important
HOT FIX COMING - this release has a bug from the latest PyTorch update on 05/13/2026, preventing ComfyUI from installing. A same day fix is coming. Look for v3.1.0-alpha_hf
This is an early alpha release of AI Playground 3.1.0 and may have minor bugs or breaks in the user experience, as it is adding new version options not previously available in AI Playground.
Note: This release's installer binary is larger than usual, containing a needed oneDNN update, scheduled to be release in PyTorch 2.12. A smaller installer binary will be released once that distribution is through PyTorch.
Version 3.1.0 alpha Updates
- Hardware Mode installer - detects hardware and provides targeted installation
- AI Playground Essentials - a curated installation of features for Intel Core Series 3 Processors
- AI Playground Studio - a named version for the standard AI Playground features targeting Intel Core Ultra Processors and Intel Arc GPUs
- AI Playground Studio, NVIDIA CUDA - Experimental installation option to run local inferencing on Nvidia GPUs via torch.cuda
- Agentic MCP Support - (for studio versions), use AI Playground Agentic chat as host MCP for 3rd party MCP Server applications
- Alternative Demo mode and features for Essentials
- Gemma 4 E4B added to Llama.cpp chat
Known Issues
- The installer can fail due to environmental issues. See the Troubleshooting guide for help
- Systems with discrete GPU and integrate GPU may need to disable iGPU when installing
- Systems with more than one GPU may need to disable idle GPUs when inferencing
- An interrupted download can block models from downloading properly. To fix this, go to the model location, delete the temporary file, and try again.
- The Colorizer workflow is limited to Xe2-supported GPUs (LNL and BMG)
- A PC restart may be needed after installation for optimal functionality and performance
- Reinstalling over an existing AI Playground installation will delete any ComfyUI nodes.
- Users can adjust settings beyond their specific system's ability resulting in failed generations
Version 3.1.0 alpha Supported Hardware
- Intel Core Series 3 Processors with 12GB+ of system memory (WCL)
- Intel Core Ultra Series 3 Processors (PTL)
- Intel Core Ultra Series 2 (H) Processors (ARL-H)
- Intel Core Ultra Series 2 (V) Processors (LNL)
- Intel Core Ultra Series 1 (H) Processors (MTL-H)
- Intel Arc B Series GPU Cards (BMG)
- Intel Arc A Series GPU Cards with 8GB+ Memory (ACM)
- Nvidia RTX GeForce GPUs
AI Playground v3.0.3 beta - hot fix
AI Playground Version v3.0.3 beta - hot fix
This is a hot fix release to AI Playground 3.0.3-beta to address bugs and version compatibility issues. For compatibility with latest OVMS, update your NPU driver to the latest version
Hot Fix Updates
- Updated version of OVMS to fix RAG embedding issues on GPU
- Updated Llama.cpp to fix issues with Qwen3-4B output as reasoning
- Removed Mistral 7B-instruct-v0.2-int4 and Phi-3-mini-4k-instruct-gq from OpenVINO model list for instability with latest backend updates
Version v3.0.3 beta Updates
See: https://github.com/intel/AI-Playground/releases/tag/v3.0.3-beta
Known Issues
- The installer can fail due to environmental issues. See the Troubleshooting guide for help
- Random black image output for SDXL (HD, ControlFace, Sketch2Photo) on Intel Core Ultra Series 3. Rerunning prompt and seed may correct. Fix will be provided in PyTorch 2.12
- Systems with discrete GPU and integrate GPU may need to disable iGPU when installing
- Systems with more than one GPU may need to disable idle GPUs when inferencing
- An interrupted download can block models from downloading properly. To fix this, go to the model location, delete the temporary file, and try again.
- The Colorizer workflow is limited to Xe2-supported GPUs (LNL and BMG)
- A PC restart may be needed after installation for optimal functionality and performance
- Reinstalling over an existing AI Playground installation will delete any ComfyUI nodes.
- Users can adjust settings beyond their specific system's ability resulting in failed generations
Version 3.0.3 beta Supported Hardware
- Intel Core Ultra Series 3 Processors (PTL)
- Intel Core Ultra Series 2 (H) Processors (ARL-H)
- Intel Core Ultra Series 2 (V) Processors (LNL)
- Intel Core Ultra Series 1 (H) Processors (MTL-H)
- Intel Arc B Series GPU Cards (BMG)
- Intel Arc A Series GPU Cards with 8GB+ Memory (ACM)
AI Playground v3.0.3 beta
AI Playground Version v3.0.3 beta
This is an incremental update to AI Playground 3.0.2-beta, with updated models, backends, feature fixes, security updates and additional mode of operations
Version v3.0.3 beta Updates
- Image Gen Manual preset now allows for 3rd party SD 1.5 and SDXL checkpoints and LoRAs
- Camera capture feature added to Chat prompt and Load Image settings tool
- Backends updated - OpenVINO: 2026.1.0, ComfyUI: 0.17.0, Llama.cpp: b8429
- Llama.cpp support for Qwen 3.5 9B
- RAG embedding enabled for NPU chat
- Chat metrics added to NPU device and models
- Updated NPU models for greater accuracy
- OpenVINO model support for GPT-OSS 20B
- New Demo Mode experience for kiosk and attendee free app demos. Enabled via settings.json
- Expanded Preset descriptions and Info icon details added to prompt settings
- Missing backend alert added when running a feature that requires a specific backend
- Input images added to history, now saved to disk, avoiding local storage issue
- Various CVE and security updates
Known Issues
- The installer can fail due to environmental issues. See the Troubleshooting guide for help
- Random black image output for SDXL (HD, ControlFace, Sketch2Photo) on Intel Core Ultra Series 3. Rerunning prompt and seed may correct. Fix will be provided in PyTorch 2.12
- Systems with discrete GPU and integrate GPU may need to disable iGPU when installing
- Systems with more than one GPU may need to disable idle GPUs when inferencing
- An interrupted download can block models from downloading properly. To fix this, go to the model location, delete the temporary file, and try again.
- The Colorizer workflow is limited to Xe2-supported GPUs (LNL and BMG)
- A PC restart may be needed after installation for optimal functionality and performance
- Reinstalling over an existing AI Playground installation will delete any ComfyUI nodes.
- Users can adjust settings beyond their specific system's ability resulting in failed generations
Version 3.0.3 beta Supported Hardware
- Intel Core Ultra Series 3 Processors (PTL)
- Intel Core Ultra Series 2 (H) Processors (ARL-H)
- Intel Core Ultra Series 2 (V) Processors (LNL)
- Intel Core Ultra Series 1 (H) Processors (MTL-H)
- Intel Arc B Series GPU Cards (BMG)
- Intel Arc A Series GPU Cards with 8GB+ Memory (ACM)
AI Playground v3.0.2 beta
AI Playground Version v3.0.2 beta
This is a Beta release to AI Playground 3.0.1-alpha, which closes on gaps and needed features in the application
Version v3.0.2 beta Updates
- Agentic chat can now tool call Image Edit presets from discussion
- Agentic chat image results show in respective mode histories
- Add Model feature in Chat fixed allowing for feature tagging
- Persistent history added for Image Gen, Image Edit and Video Gen modes
- Image Histories now hold "Before" reference images used for Image Gen or Image Edit
- Pro Preset upgraded to Flux.2 Klein model
- Control Face Preset now includes a "Quality" variation using Z-Image-Turbo model
- Image Edit results include "Post to Image Edit mode" icon tool
- SafetyCheck added to Image Edit presets that generate additional content
- SafetyChecker offline timeout bug fixed
- Wan 2.1 VACE bug fixed on dGPUs - (reserve vRAM added back to ComfyUI Parameters)
- HuggingFace Mirror URL option added to App Settings
- ComfyUI start up parameter added to the ComfyUI backend settings options
- ComfyUI install now update front end to current version
- Persistent Developer Console Mode added to app settings
- Keep Models Loaded option added to app settings
Known Issues
- The installer can fail due to environmental issues. See the Troubleshooting guide for help
- Random black image output for SDXL (HD, ControlFace, Sketch2Photo) on Intel Core Ultra Series 3 with latest driver. Rerunning prompt and seed may correct.
- Gemma-3-4B vision issues spotted on earlier gen and lower memory Intel Core Ultra Systems
- Systems with discrete GPU and integrate GPU may need to disable iGPU when installing
- Systems with more than one GPU may need to disable idle GPUs when inferencing
- An interrupted download can block models from downloading properly. To fix this, go to the model location, delete the temporary file, and try again.
- The Colorizer workflow is limited to Xe2-supported GPUs (LNL and BMG)
- A PC restart may be needed after installation for optimal functionality and performance
- Reinstalling over an existing AI Playground installation will delete any ComfyUI nodes.
- Users can adjust settings beyond their specific system's ability resulting in failed generations
Version 3.0.2 beta Supported Hardware
- Intel Core Ultra Series 3 Processors (PTL)
- Intel Core Ultra Series 2 (H) Processors (ARL-H)
- Intel Core Ultra Series 2 (V) Processors (LNL)
- Intel Core Ultra Series 1 (H) Processors (MTL-H)
- Intel Arc B Series GPU Cards (BMG)
- Intel Arc A Series GPU Cards with 8GB+ Memory (ACM)