Open Creative Studio 13.0 introduces a staggering number of new features and a radical redesign of some of its core functions.
New Functions and Features
OCS now features a new Lens Effect function in the Image Finishers section of the Image Enhancement pipeline.
With it, you can apply effects like lens distortion, chromatic aberration, or vignette. The lens distortion effect can also be applied in reverse to correct the natural distortion occurring in smartphone images.
OCS features a new and simplified Image Saver node, part of the OCS Nodes node suite.
By default, it’s configured to save the OCS log in the EXIF UserComment metadata tag of all supported image formats (.png, .jpeg/jpg, .webp) so you can read it inside any image manipulation software that supports EXIF data.
Just like the previous Image Saver node, this new node also allows you to, optionally, save the entire workflow in the Extra metadata section of .png images.
Every function in OCS with a resizing capability has been redesigned to use a new all-in-one Image Resize node with advanced capabilities:
- Image Resizer (part of the Image Finishers functions)
- Source Resizer for I2V/V2V (part of the Video Manipulators function)
- Upscaler (Supir)
Every image generation pipeline (FLUX.1, SD1.5, SDXL, SD3.5), the RePainter and InPainter functions, the Object Swapper function, and the Face and Hand Refiner functions now use new nodes to load their diffusion models and their VAE models.
These new nodes offer granular capability to define weight and compute dtypes, allowing you to use OCS with resource-constrained computers in a much easier way.
These new nodes are also configured to automatically take advantage of Sage Attention if your machine is configured to use it.
OCS features a new always-on Attention Mode function, which centralizes the configuration of the attention mode for all video models.
The default configuration is set to Scaled Dot Product Attention (sdpa), but you can change it to Sage Attention (sageattn) or Flash Attention (flash_attn), depending on how your machine is configured.
Redesigned Functions and Features
The Repainter (I2I) function now depends on the new FLUX.1 Kontext Dev model.
This extremely versatile model allows you to repaint source images, inpaint specific details, recolor, swap objects, merge multiple subjects, and more.
This extremely versatile model allows you to fully repaint source images, inpaint specific details, recolor old photos, swap objects, merge multiple subjects from different source images, transfer style, and more.
To facilitate some of these use cases, the Prompt for Repainter (I2I) function has been redesigned to feature a number of predefined prompts.
For now, the existing OCS functions dedicated to these use cases will remain. With time, as we understand the quality and prompt limitations of FLUX.1 Kontext Dev, we’ll revisit the decision.
OCS features a completely redesigned Prompt for T2I function.
Redesigned Sections
OCS features a completely redesigned Control Panel section, where the switches are organized by focus: text generation, image generation, video generation, and audio generation.
The rest of the Control Panel section now features Formats & Specifications, where you define the formats of your images, audio, and videos, and a series of Loaders and Savers, where you define what images to upload, and you can see the outcome of your generations.
Every prompt used throughout OCS has been moved inside a new Text Generation pipeline. Here you’ll be able to set independent prompts for image generation (T2V and I2I), image upscaling, video generation (T2V and V2V), and audio generation (speech, audio tags, and lyrics).
In the future, this new pipeline will host new functions to manipulate text for content production from a source input like a book or an RSS feed.
The Text Generation pipeline now features a number of new Prompt Enricher and Prompt Generator functions.
Over time, the system prompts used by these functions will be refined and improved.
The Image Enhancement section of OCS has been redesigned to have completely separated RePainting and InPainting pipelines.
[EA5] The Caption Generation function has been renamed Caption Generator for IPAdapter/Redux/I2V.
The Positive Prompt Builder function has been renamed to Prompt for T2I.
[EA5] APW Nodes snapshot now includes LoRA Manager. The node suite is not essential for the functioning of APW, but it provides a mature LoRA catalog functionality.
[EA4] APW now includes a new Audio Generation Pipeline featuring text-to-audio (T2A) generation via the remarkable ACE Step model and its LoRAs.
Use the Positive Prompt Builder function to input your lyrics and your audio tags (which explain the model what type of sound you are trying to produce).
This is just a first step. In the future, APW will be able to pair audio with video.
[EA4] The Image Generation Pipeline section of APW now supports a new model: OpenAI GPT-Image-1 (GI1). You can generate images with GI1 and then process them in the Image Enhancement Pipeline, for example to repaint them with a FLUX.1 LoRA.
Notice that, for the time being, this node requires an API key generated by a verified OpenAI account. If you already have an API key, but you are not a verified organization, this function will fail.
[EA4] The Image Spec and Video Spec functions are now merged into a single function: Image/Video Spec. This new function features three new APW Nodes, Image Size (Local), Image Size (Cloud), Video Size (Local), which organize the target resolution of your images in a way that is more suitable for APW.
[EA4] The Master Log function now also displays in the terminal the name of each LoRA used by each image generation model.
[EA4] Every APW function using the Florence-2 model (Caption Generator, Upscaler) is now configured to automatically convert its weights to the safetensors format for faster loading.
[EA4] The Image/Video Uploader function has been renamed into Image/Audio/Video Uploader and now features a dedicated Source Audio section. It does nothing useful for now, but it will be needed in the future.
[EA3] APW now supports the new WanVideo 2.1 FLF model which allows you to define the first and last frame of your videos. Set your start frame as 1st Reference Image and your last frame as 2nd Reference Image in the Image/Video Uploader function.
Notice that, for now, the 1st Reference Image and the 2nd Reference Image are will not be automatically resized to match the Video Spec function settings. You need to do it manually.
[EA3] The Video Generation Pipeline has been redesigned to accommodate up to 4 video modalities for each supported video generation model.
[EA3] APW now allow you to generate images with multiple AI models at the same time (FLUX.1, SD1.5, SDXL, and SD3.5) and process their output in the Image Enhancement Pipeline sequentially. This means, for example, that you can upscale all four images with SUPIR in one single operation.
If no image generator is active, an image uploaded via the Image/Video Uploader function will be used as fallback. This allows you to use Image Enhancement Pipeline functions to improve pre-existing images.
[EA3] The Image Enhancement Pipeline now features a new function: Image Resizer. For now, it will resize any image the dimensions specified in the Video Spec function. This is useful to automatically format images before I2V generation.
[EA3] The Video Generation Pipeline now features a new function: Source Video Resizer. It will resize every frame of a video uploaded via the Image/Video Uploaded function to match the dimensions specified in the Video Spec function. This is useful to automatically format frames before V2V generation.
[EA3] The FLUX Configurator function has been reconfigured to increase quality and crispness of images generated with FLUX.1 Dev.
[EA3] The Image Chooser function now features @chrisgoringe’s new Image Filter custom node suite in place of his old Image Picker, now deprecated.
[EA3] The Final Image Saver function is now split in two: Final Image Saver and Final Image Saver (w/o) Metadata. The latter is now optional and can be activated/deactivated in the Extras section of the Controller function.
[EA3] The Image Comparer function now features the capability to also compare the image you just generated with the last image you generated, automatically loaded from your Output folder.
[EA3] The FLUX ControlNet functions Canny, Depth, and Pose have been reconfigured to best support the new FLUX.1-dev-ControlNet-Union-Pro-2.0 released by Shakker Labs.
[EA3] The Face Swapper function now uses the SFW version of the Reactor custom node suite. The previous version of this suite is not available anymore on GitHub.
[EA2] APW now supports CFG Zero Star for FLUX, SD3.5L, and WanVideo (in all video modalities). CFG Zero Star is enabled by default for both FLUX And SD3.5L.
[EA1] APW now supports the WanVideo 2.1 model for I2V, V2V, and T2V generation.
[EA1] APW now supports the Hunyuan Video for I2V generation.
[EA1] APW now supports the CogVideoX for T2V generation.
[EA1] APW 13.0 introduces an almost complete redesign:
The Image Generation Pipeline now features separate function for SDXL and SD1.5. No more need to change dozens of settings to switch between SDXL and SD1.5.
Each subsection in the Image Generation Pipeline now has ample space to configure and organize your LoRAs. LoRAs for SDXL and SD1.5 are no more second class citizens.
The design of each Video Generation Pipeline subsection has been redesigned to be consistent across video models.
The Front End section now lives just below the Control Panel section, for easier access.
The Intermediate Images Saver function has been greatly expanded, and it’s now moved to the Aux Functions section of APW.
A new Video Spec function has been added to the Control Panel section, so you can centrally set the size and the FPS of the videos to be generated.
Most of the other functions have been aggregated in a new Image Enhancement Pipeline.
[EA1] The Latent function has been renamed to Image Spec.
[EA1] The Training Helper function has been moved back next to the Caption Generator function.
Removed Functions and Features
The Prompt Tagger function has been temporarily removed while it’s being redesigned to work for each image generation pipeline independently.
The XYZ Plot function has been removed because the custom node suite it depends on has not been updated for months. The absence of this function greatly speeds up the loading of OCS.
A stand-alone alternative is in the works.