Version History
Here you can find the changelog of Sogni - AI Art Generator since it was posted on our website on 2023-07-05.
The latest version is 2.5 and it was updated on soft112.com on 22 April, 2024.
See below the changes in each version:
version 2.5
posted on 2024-04-03
Apr 3, 2024
Version 2.5
• New Sogni Canvas: Extend your images from the standard 2048x2048 pixels to an impressive 4096 pixels in any direction you choose, enabling custom aspect ratios while enriching your images with intricate details, rather than simply cropping them.
• Enhanced download reliability for Stable Diffusion, ControlNet, and preprocessor models, ensuring accurate file handling and storage.
• Significant optimization enhancements.
• Fixes a bug that previously led to the accumulation of unnecessary temporary files.
• Enables saving your canvas as a project file.
• Introduces new UI refinements and practical detailing.
• Utilize a pinch gesture to easily switch between zoom options: Actual Size and Fit to Window.
• All models supporting ControlNet have been updated for enhanced coherence.
• New models:
Anything XL
Anything XL 896px [8bit]
EdgeOfRealism v2
EdgeOfRealism v2 [6bit]
EpiCRealism pure Evolution v5
EpiCRealism pure Evolution v5 [6bit]
Sweet-mix v2.2
Sweet-mix v2.2 [6bit]
YesMix v4
YesMix v4 [6bit]
LCM - Comic Craft b4
LCM - Comic Craft b4 [6bit]
LCM - CyberRealistic v4.2
LCM - CyberRealistic v4.2 [6bit]
LCM - DreamShaper v8
LCM - DreamShaper v8 [6bit]
version 2.1.2
posted on 2024-03-18
Mar 18, 2024
Version 2.1.2
• The updated ControlNet Segmentation preprocessing model delivers better results in removing the background or subject from an image.
• New keyboard shortcuts: Use the arrow keys to navigate between batch-generated images.
• New realistic SD Models:
RealDream v12
RealDream v12 [6bit]
RealDream Directors Cut v2
RealDream Directors Cut v2 [6bit]
• Minor bug fixes and optimization improvements.
version 2.1.1
posted on 2024-02-27
Feb 27, 2024
Version 2.1.1
• Support for SDXL-Lightning models.
• Quick-preview for images is now available while batch is still in progress.
• New keyboard shortcuts:
Use 'esc' to exit the lasso tool or individual image viewing and return to the batch grid. Use alt-command-G to add/remove items to the Gallery.
• Fixes a bug that could cause animation glitches in the Controls Bar.
• The Safe Content Filter is now a bit stricter.
• Various UI refinements.
• Other bug fixes and optimization enhancements.
• New Sogni models featuring enhanced coherence with ControlNet Models:
Sogni Artist v1.0
Sogni Artist v1.0 [6bit]
Sogni Photo v1.0
Sogni Photo v1.0 [6bit]
LCM - Sogni v1.0
LCM - Sogni v1.0 [6bit]
• New variant models:
SDXL-Lightning - JuggernautXL 9 + RD Photo2
SDXL-Lightning - JuggernautXL 9 + RD Photo2 896x896 [8bit]
SDXL-Turbo - DreamShaper v2.1
SDXL-Turbo - DreamShaper v2.1 896x896 [8bit]
LCM - CyberRealistic v3.3
LCM - CyberRealistic v3.3 [6bit]
CyberRealistic Classic v3.1
CyberRealistic Classic v3.1 [6bit]
LEOSAM's HelloWorld XL v5 GPT4V
LEOSAM's HelloWorld XL v5 GPT4V 896px [8bit]
ZavyChromaXL v4
ZavyChromaXL v4 896px [8bit]
SDXL Yamer's Realistic 5 + RunDiffusion
SDXL Yamer's Realistic 5 + RunDiffusion 896px [8bit]
version 2.1
posted on 2024-02-20
Feb 20, 2024
Version 2.1
. Camera Roll controls for Guide Image and ControlNet. Now you can set and animate the cameras in all axes: zoom, pan, and roll.
. Gallery: Select and delete multiple items at once.
. Gallery: Full-screen-sized image quick preview.
. Bug fixes and UI refinements.
version 2.0.5
posted on 2024-02-14
Feb 14, 2024
Version 2.0.5
. Fixes a bug that affected saving/opening projects containing images with transparency, resulting in images having a white background upon loading the project file.
. Provides access to the Sogni Discord from all help views.
. Includes UI refinements and other bug fixes.
In case you missed the previous update:
. You can now undo and change the upscaling method when working with individual images.
. Introducing 30 new style presets for illustrations, paintings, and crafts.
. New photorealistic models:
NewRealityXL Photographic v2.1
NewRealityXL Photographic v2.1 [8bit]
NewRealityXL Photographic v2.1 896px [8bit]
version 2.0.4
posted on 2024-02-07
Feb 7, 2024
Version 2.0.4
. You can now undo and change the upscaling method when working with individual images.
. Introducing 30 new style presets for illustrations, paintings, and crafts.
. New photorealistic models:
NewRealityXL Photographic v2.1
NewRealityXL Photographic v2.1 [8bit]
NewRealityXL Photographic v2.1 896px [8bit]
. UI refinements, bug fixes, and optimization enhancements.
If you missed the previous updates:
. New Timeline Keyframes Module: Use keyframes to control the progress of generation settings across the animation's timeline, animate camera movements, and more.
. Save, edit, and share your projects as '.sogni' files, enabling you to edit previous creations, collaborate on the same project with other creatives, and keep a record of the creative process involved in the generation of your images and animations.
. Updated Body-Pose capture preprocessor provides more accurate skeleton rendering.
. Filter the SD models list by style tendency or type to make it easier to find the model you are looking for.
. UI optimization results in smoother controls.
. New models:
SDXL-Turbo - Pixel Art
SDXL-Turbo - Pixel Art [8bit]
SDXL-Turbo - Pixel Art 896x896 [8bit]
Playground v2 - 1024px Aesthetic Model
Playground v2 - 1024px Aesthetic Model [8bit]
Playground v2 - 896px Aesthetic Model [8bit]
. UI refinements.
. Help content updates.
version 2.0.3
posted on 2024-01-30
Jan 30, 2024
Version 2.0.3
. New Timeline Keyframes Module: Use keyframes to control the progress of generation settings across the animation's timeline, animate camera movements, and more.
. Save, edit, and share your projects as '.sogni' files, enabling you to edit previous creations, collaborate on the same project with other creatives, and keep a record of the creative process involved in the generation of your images and animations.
. Updated Body-Pose capture preprocessor provides more accurate skeleton rendering.
. Filter the SD models list by style tendency or type to make it easier to find the model you are looking for.
. UI optimization results in smoother controls.
. New models:
SDXL-Turbo - Pixel Art
SDXL-Turbo - Pixel Art [8bit]
SDXL-Turbo - Pixel Art 896x896 [8bit]
Playground v2 - 1024px Aesthetic Model
Playground v2 - 1024px Aesthetic Model [8bit]
Playground v2 - 896px Aesthetic Model [8bit]
. UI refinements.
. Important bug-fixes.
. Addresses minor bugs that were introduced in v2.0 - 2.0.2.
. Help content updates.
version 2.0.1
posted on 2024-01-26
Jan 26, 2024
Version 2.0.1
. New Timeline Keyframes Module: Use keyframes to control the progress of generation settings across the animation's timeline, animate camera movements, and more.
. Save, edit, and share your projects as '.sogni' files, enabling you to edit previous creations, collaborate on the same project with other creatives, and keep a record of the creative process involved in the generation of your images and animations.
. Updated Body-Pose capture preprocessor provides more accurate skeleton rendering.
. Filter the SD models list by style tendency or type to make it easier to find the model you are looking for.
. UI optimization results in smoother controls.
. UI refinements.
. Important bug-fixes.
. Help content updates.
version 1.8.7
posted on 2024-01-02
Jan 2, 2024
Version 1.8.7
. Resolves bugs that could affect the download and installation of ControlNet models and preprocessors.
. Resolves a bug that could keep the upscaling model enabled after opting to deactivate it from the Real-Time Generation popover message.
If you missed the previous update:
. Share groups of images and videos from the Gallery.
. New Models:
SDXL-Turbo - TurboVisionXL v4.3.1
SDXL-Turbo - TurboVisionXL v4.3.1 896x896 [8bit]
A-Zovya RPG Artist Tools v4
A-Zovya RPG Artist Tools v4 [6bit]
Architecture Ra-render v3
Architecture Ra-render v3 [6bit]
. Correctly sized app icons.
version 1.8.5
posted on 2023-12-19
Dec 19, 2023
Version 1.8.5
. A new collection of Stable Diffusion models by Sogni, designed to generate higher quality images in shorter timeframes.
Sogni Artist v.Alpha3
Sogni Artist v.Alpha3 [6bit] (New default model)
Sogni Photo v.Alpha2
Sogni Photo v.Alpha2 [6bit]
Sogni.XL v.Alpha2
Sogni.XL v.Alpha2 [8bit]
Sogni.XL v.Alpha2 896x896 [8bit]
LCM - Sogni v.Alpha1
LCM - Sogni v.Alpha1 [6bit]
SDXL-Turbo - Sogni.XLT v.Alpha1
SDXL-Turbo - Sogni.XLT v.Alpha1 [8bit]
SDXL-Turbo - Sogni.XLT v.Alpha1 896x896 [8bit]
. Bug fixes.
. UI-Refinements.
. Help content updates
If you missed the previous updates:
. New "Lasso Eraser / Mask Tool": Use this tool to create masks from your images, and combine them with the InPaint ControlNet model to fill the cleared areas with new content. This is a really powerful solution when you wish to quickly remove objects from an image, change background, subjects, faces, etc.
. The Stable Diffusion model import feature now supports SDXL-Turbo models.
. Batch Generation: Request multiple image variations in one go. Generate a spectrum of proposals from the same settings, empowering you to choose the options that perfectly align with your vision and requirements.
. Add individual images from a batch to the Gallery or the complete batch as a group.
version 1.8.1
posted on 2023-12-16
Dec 16, 2023
Version 1.8.1
. New "Lasso Eraser / Mask Tool": Use this tool to create masks from your images, and combine them with the InPaint ControlNet model to fill the cleared areas with new content. This is a really powerful solution when you wish to quickly remove objects from an image, change background, subjects, faces, etc.
. The Stable Diffusion model import feature now supports SDXL-Turbo models.
. New SD model: LCM - Stable Yogi's Realistic.
. Help content updates.
. Bug-fixes.
. UI refinements.
If you missed the previous updates:
. Batch Generation: Request multiple image variations in one go. Generate a spectrum of proposals from the same settings, empowering you to choose the options that perfectly align with your vision and requirements.
. Add individual images from a batch to the Gallery or the complete batch as a group.
. New Models:
SDXL-Turbo - PLUS-RED TEAM [8bit]
SDXL-Turbo - PLUS-RED TEAM 768px [8bit]
LCM - DreamCrafter v1
LCM - DreamCrafter v1 [6bit]
version 1.8
posted on 2023-12-12
Dec 12, 2023
Version 1.8
. Batch Generation: Request multiple image variations in one go. Generate a spectrum of proposals from the same settings, empowering you to choose the options that perfectly align with your vision and requirements.
. Add individual images from a batch to the Gallery or the complete batch as a group.
. New Models:
SDXL-Turbo - PLUS-RED TEAM [8bit]
SDXL-Turbo - PLUS-RED TEAM 768px [8bit]
LCM - DreamCrafter v1
LCM - DreamCrafter v1 [6bit]
. Updated style presets.
. Bug fixes.
. Optimization.
. UI refinements.
. Help content updates.
If you missed the previous updates:
. Real-time image generation: When working with LCM or SDXL-Turbo and the LCM Scheduler, the option to generate images in "Real-Time" as you make changes to your settings becomes available. This feature allows you to see result changes as you type your prompt, adjust the Guidance Scale, switch styles, etc.
. Support for SDXL-Turbo: Generate SDXL-quality images in just 1-2 steps.
. New SDXL-Turbo models:
SDXL-Turbo v1
SDXL-Turbo v1 [8bit]
SDXL-Turbo v1 512x512 [8bit]
SDXL-Turbo - Realities Edge XL
SDXL-Turbo - Realities Edge XL 512x512 [8bit]
SDXL-Turbo - Tertium v1
SDXL-Turbo - Tertium v1 512x512 [8bit]
SDXL-Turbo - TurboVisionXL Alpha
SDXL-Turbo - TurboVisionXL Alpha [8bit]
SDXL-Turbo - TurboVisionXL Alpha 512x512 [8bit]
SDXL-Turbo - Unstable Diffusers - YamerMIX v10 [8bit]
SDXL-Turbo - Unstable Diffusers - YamerMIX v10 512x512 [8bit]
SDXL-Turbo - Yamer's Perfect Design v5.SpecialGift [8bit]
SDXL-Turbo - Yamer's Perfect Design v5.SpecialGift 768px [8bit]
(These models are currently released by Stability AI under a non-commercial research license and permit personal, non-commercial use only.)
version 1.7.1
posted on 2023-12-06
Dec 6, 2023
Version 1.7.1
. Real-time image generation: When working with LCM or SDXL-Turbo and the LCM Scheduler, the option to generate images in "Real-Time" as you make changes to your settings becomes available. This feature allows you to see result changes as you type your prompt, adjust the Guidance Scale, switch styles, etc.
. Resolves a bug that erroneously listed SDXL-Turbo models as available with the Neural Engine, leading to a download error
. Support for SDXL-Turbo: Generate SDXL-quality images in just 1-2 steps.
. New SDXL-Turbo models:
SDXL-Turbo v1
SDXL-Turbo v1 [8bit]
SDXL-Turbo v1 512x512 [8bit]
SDXL-Turbo - Realities Edge XL
SDXL-Turbo - Realities Edge XL 512x512 [8bit]
SDXL-Turbo - Tertium v1
SDXL-Turbo - Tertium v1 512x512 [8bit]
SDXL-Turbo - TurboVisionXL Alpha
SDXL-Turbo - TurboVisionXL Alpha [8bit]
SDXL-Turbo - TurboVisionXL Alpha 512x512 [8bit]
SDXL-Turbo - Unstable Diffusers - YamerMIX v10 [8bit]
SDXL-Turbo - Unstable Diffusers - YamerMIX v10 512x512 [8bit]
SDXL-Turbo - Yamer's Perfect Design v5.SpecialGift [8bit]
SDXL-Turbo - Yamer's Perfect Design v5.SpecialGift 768px [8bit]
(These models are currently released by Stability AI under a non-commercial research license and permit personal, non-commercial use only.)
. Help content updates.
version 1.7
posted on 2023-12-02
Dec 2, 2023
Version 1.7
. Real-time image generation: When working with LCM or SDXL-Turbo and the LCM Scheduler, the option to generate images in "Real-Time" as you make changes to your settings becomes available. This feature allows you to see result changes as you type your prompt, adjust the Guidance Scale, switch styles, etc.
. Support for SDXL-Turbo: Generate SDXL-quality images in just 1-2 steps.
. New SDXL-Turbo models:
SDXL-Turbo v1
SDXL-Turbo v1 [8bit]
SDXL-Turbo v1 512x512 [8bit]
SDXL-Turbo - Realities Edge XL
SDXL-Turbo - Realities Edge XL 512x512 [8bit]
SDXL-Turbo - TurboVisionXL Alpha
SDXL-Turbo - TurboVisionXL Alpha [8bit]
SDXL-Turbo - TurboVisionXL Alpha 512x512 [8bit]
(These models are currently released by Stability AI under a non-commercial research license and permit personal, non-commercial use only.)
. Critical bug fix: Resolves a bug that could cause the app to crash when unsetting ControlNet.
. Help content updates.
. Optimization and UI refinements.
version 1.5.7
posted on 2023-11-20
Nov 20, 2023
Version 1.5.7
. Latent Consistency Models (LCM): Generate images in just 1-3 seconds and animation frames in 1 second using these models, which only require 2-8 steps to produce high-quality images.
. Three new LCM models: MagicMix + Realistic Vision v7, Sote-Mix, Stable Diffusion v1.5 Base. All three support Guide Image and ControlNet.
. Three new SDXL models: Blue Pencil XL v0.8, Juggernaut XL Version 6 + RunDiffusion, Starlight XL Animated v3.
. Schedulers and Time Spacing options: Experiment with a growing selection of schedulers and time-spacing algorithms to find the combination that works best for your creations.
. Scheduler options: DPM Solver Multistep, PNDM (Pseudo-linear multi-step), LCM (Latent Consistency Models).
. Time Spacing options: Linear, Leading, Karras.
. Bug fixes.
. Optimization.
. UI refinements.
If you missed the recent updates:
. Keep your Gallery and projects organized by grouping items into albums. Drag and drop an item onto another one to create a group or use the multiple selection tool.
. Support for models with an 896x896 output size.
. Import SDXL models with custom sizes (768, 896, 1024).
. The new Upscale button on the Toolbar allows you to turn OFF automatic upscaling and decide later, enabling better performance, especially during animation generation.
. Overall performance optimization.
. Improved performance in image and video upscaling.
. Enhanced performance in ControlNet preprocessors (Up to 4 times faster).
. New "Optimize Memory Usage" setting for devices with limited RAM. Enabling this setting can be advantageous when running resource-intensive models on systems with less than 32GB of RAM, or if the generation process is causing your Mac to run slowly.
. Import Stable Diffusion (CoreML) Models. You can import your own trained models or download them from sources like HuggingFace, GitHub, CivitAI, etc.
. New Segmentation preprocessor for ControlNet, including subject and background masks. Use it with the segmentation model to capture scene composition and with the InPaint model to change the background or subjects only.
version 1.5.6
posted on 2023-11-13
Nov 13, 2023
Version 1.5.6
. Keep your Gallery and projects organized by grouping items into albums. Drag and drop an item onto another one to create a group or use the multiple selection tool.
. Support for models with an 896x896 output size.
. Import SDXL models with custom sizes (768, 896, 1024).
. Bug fixes.
. UI refinements.
If you missed the recent updates:
. The new Upscale button on the Toolbar allows you to turn OFF automatic upscaling and decide later, enabling better performance, especially during animation generation.
. Overall performance optimization.
. Improved performance in image and video upscaling.
. Enhanced performance in ControlNet preprocessors (Up to 4 times faster).
. New "Optimize Memory Usage" setting for devices with limited RAM. Enabling this setting can be advantageous when running resource-intensive models on systems with less than 32GB of RAM, or if the generation process is causing your Mac to run slowly.
. Import Stable Diffusion (CoreML) Models. You can import your own trained models or download them from sources like HuggingFace, GitHub, CivitAI, etc.
. New Segmentation preprocessor for ControlNet, including subject and background masks. Use it with the segmentation model to capture scene composition and with the InPaint model to change the background or subjects only.
version 1.5.3
posted on 2023-11-05
Nov 5, 2023
Version 1.5.3
. Import Stable Diffusion (CoreML) Models. You can import your own trained models or download them from sources like HuggingFace, GitHub, CivitAI, etc.
. New models: DynaVisionXL v0.5.5.7, AnimeLineart LoRA + Anything V5Ink, KidsIllustration LoRA + Lyriel, kLabyrinth LoRA, FairyTaleAI LoRA + ReVAnimated.
. Optimization improvements.
. UI refinements.
. Important Bug fixes.
If you missed the recent updates:
. New Segmentation preprocessor for ControlNet, including subject and background masks. Use it with the segmentation model to capture scene composition and with the InPaint model to change the background or subjects only.
. New animation module: Create videos at 2048x2048px, animate anything you can dream of in any way you can imagine.
. Pause the generation process and preview a video of your animation.
. Text-to-video, image-to-video and video-to-video.
. Set videos as Guide Image and/or ControlNet reference image.
. Automatic Guide Image and ControlNet images for animations.
. Save videos and frame images folders to the gallery.
. Restore a video from the gallery and generate more frames to extend it.
. Change generation settings live during the animation generation process.
. New Camera zoom & pan controls.
version 1.2.4
posted on 2023-10-15
Oct 15, 2023
Version 1.2.4
This update brings important bug fixes and is recommended for all users:
. Fixes a critical bug that could cause the app to crash when a Guide Image with a high strength value and a low number of steps is set.
. Turn seed shuffle on or off without the need to expand the seed module.
If you missed the recent updates:
. Capture human faces and poses with the new ControlNet preprocessors.
. Generate pose-skeleton and face landmarks reference images for OpenPose.
. New Camera Zoom & Pan tool for Guide Image and ControlNet reference images.
. Link Guide Image and ControlNet camera settings.
. Zoomed-out images retain a transparent background when used with the InPaint model.
. First and last page navigation buttons in the Gallery.
version 1.2.1
posted on 2023-10-11
Oct 11, 2023
Version 1.2.1
This update brings significant improvements and is recommended for all users:
. Achieve ultra realistic results with the EpiCRealism pure Evolution v3 model.
. New tools to rotate and flip your images.
. Bug fixes.
. UI refinements.
If you missed the recent updates:
. The new ControlNet module allows you to transfer compositions or human poses from a reference image into a newly generated image, giving you control over various aspects such as the placement of subjects, their appearance, poses, environment, background, objects, lighting, and more.
. There are 14 different ControlNet models available, including Canny, Depth, InPaint, OpenPose, Instruct Pixel2Pixel, Scribble, and others.
. ControlNet preprocessors include Depth, Sketch/Outline, and Invert Colors.
. Now all SD models support Guide Image, except for SDXL.
version 1.1.2
posted on 2023-09-30
Sep 30, 2023
Version 1.1.2
This update brings significant improvements and is recommended for all users:
. Important bug fixes.
. SDXL 1.0 with refiner, SDXL 1.0 4-bit, and GhostMix v2 models are now available.
. The list of models now indicates which ones support Guide Images (text guided image-to-image)
. Buttons to Cancel/Reload a model.
. Gallery UI refinements.
. Other UI adjustments.
If you missed the recent updates:
. New image-to-image capabilities allow you to create variations of your generated images.
. Import any image and use it as Guide Image.
. Zoom-in to view the generated image at its actual size.
. Fixes a critical bug that affected the Gallery when filtering favorites.
. Optimized for macOS 14.
. Run Stable Diffusion XL 1.0 (macOS 14 only).
. New, lighter 6-Bit versions of most models make it possible for less powerful Macs to run Sogni (macOS 14 only).
. Additional CoreML models: Openjourney v4, DreamShaper v8, DreamShaper XL v1.0, RealisticVision v5.1, and Sweet-mix v2.1
. Improved memory management.
. Improved Gallery performance.
. Enhancements to the Safe Content Filter make it more accurate and safer.
version 1.01
posted on 2023-09-26
Sep 26, 2023
Version 1.01
This update brings significant improvements and is recommended for all users:
. New image-to-image capabilities allow you to create variations of your generated images.
. Import any image and use it as Guide Image.
. Zoom-in to view the generated image at its actual size.
. Fixes a critical bug that affected the Gallery when filtering favorites.
. UI adjustments
. Other bug fixes and improvements.
If you missed the previous update:
. Optimized for macOS 14.
. Run Stable Diffusion XL 1.0 (macOS 14 only).
. New, lighter 6-Bit versions of most models make it possible for less powerful Macs to run Sogni (macOS 14 only).
. Additional CoreML models: Openjourney v4, DreamShaper v8, DreamShaper XL v1.0, RealisticVision v5.1, and Sweet-mix v2.1
. Improved memory management.
. Improved Gallery performance.
. The Gallery search field makes it easier to find your images.
. When restoring an image from the Gallery, the model and processing units that were used to generate it are now included as part of the image settings.
. Censored images that were generated with the Safe Content Filter disabled and added to the gallery are now concealed when the filter is turned ON.
. Enhancements to the Safe Content Filter make it more accurate and safer.
version 1.0
posted on 2023-09-18
Sep 18, 2023
Version 1.0
This update brings significant improvements in performance and image generation quality. It is recommended for all users:
. Optimized for macOS 14.
. Run Stable Diffusion XL 1.0 (macOS 14 only).
. New, lighter 6-Bit versions of most models make it possible for less powerful Macs to run Sogni (macOS 14 only).
. Additional CoreML models: Openjourney v4, DreamShaper v8, DreamShaper XL v1.0, RealisticVision v5.1, and Sweet-mix v2.1
. Improved memory management.
. Improved Gallery performance.
. The Gallery search field makes it easier to find your images.
. When restoring an image from the Gallery, the model and processing units that were used to generate it are now included as part of the image settings.
. Censored images that were generated with the Safe Content Filter disabled and added to the gallery are now concealed when the filter is turned ON.
. Enhancements to the Safe Content Filter make it more accurate and safer.
. UI adjustments.
. Other bug fixes and improvements.
version 0.8
posted on 2023-08-26
Aug 26, 2023
Version 0.8
. We have revised and optimized all style presets to achieve better generation results.
If you missed the previous update:
We made significant improvements in terms of speed and image generation quality.
. Reduced session initiation times
. Accelerated image generation
. Other performance optimizations
. Choose the number of previews to get a glimpse of the generation progress
. Implementation of Control-net, providing greater depth in images
. Fixes a major bug caused by using overly large seeds that the model can't handle. This is done by checking the inputs in the seed field and making sure that randomly generated seeds are within an acceptable range.
. Addresses bugs that could lead to issues when adding or removing images from the Gallery
. Resolves a bug that impacted the functionality of the "Save as..." command
. Added prompt maximum length indicator
. UI enhancements
version 0.5
posted on 2023-08-17
Aug 17, 2023
Version 0.5
We've made significant improvements in terms of speed and image generation quality. This update is strongly recommended for all users.
. Reduced session initiation times
. Accelerated image generation
. Choose the number of previews to get a glimpse of the generation progress
. Implementation of Control-net, providing greater depth in images
. UI enhancements
. Added prompt maximum length indicator
. New style presets for Disco diffusion, Funko Pop, Inkpunk, Openjourney, Paper cut, Waifu; also, adjustments to other presets
. Other performance optimizations
. Minor bug fixes
version 0.4
posted on 2023-08-09
Aug 9, 2023
Version 0.4
. Device passcode or TouchID is now required to turn OFF the Safe Content Filter.
. New psychedelic style preset.
. SD Models are now listed with shorter and friendlier names.
. Enhanced Help contents.
. Added link to Terms & Conditions from the Help menu.
. UI refinements.
version 0.3
posted on 2023-07-30
Jul 30, 2023
Version 0.3
. Fixes a bug that could duplicate the negative prompt when restoring images from the gallery.
. Introduces a new default negative prompt to help you generate better images.
. Performance enhancements.
. UI adjustments.
. Other minor bug fixes and optimizations.
version -
posted on 2023-07-14
version 0.2
posted on 2023-07-14
Jul 14, 2023
Version 0.2
. UI refinements
. Enhanced error messages with more details
. Optimization and minor bug fixes