Welcome to my 💫🌄 Friendly Ideogram 4
✨ Less mess, more magic and control
Feel like a real designer with Ideogram 4 and my friendly workflow!
I offer my workflow with LLM Enhancer & Prompt Builder modes. Compose each part of your generation!
The latest version of ComfyUI is required.
💻 System requirements:
Minimum system requirements:
RTX 3000-s, 8GB+ VRAM, 32GB+ RAM, 8-core processor, SSD, latest ComfyUI
📌 Detailed tips and links to models in the workflow
✨ Workflow features:
Extremely user-friendly interface
Maximum performance even on low-end systems with at least 8GB of VRAM
LLM Enhancer and Prompt Builder implementation
Support for up to 2 Lora
Detailed tooltips and links to models
Manual random seed for complete control over generations
Model overview
Ideogram 4.0 is a powerful, open-weight text-to-image model designed for professional-grade visual creation. It features a 9.3 billion parameter single-stream Diffusion Transformer (DiT) architecture that was trained entirely from scratch
Key Capabilities & Features
Production-Grade Typography: It offers crystal-clear, dense text rendering across multiple languages directly inside the composition, rather than as a fragile overlay
Native Transparency: The model can generate images with native transparent backgrounds, eliminating the need for post-processing background removal
High Resolution & Flexibility: It supports native 2K resolution output and flexible aspect ratios up to 2048 pixels on each side
Advanced Controllability: The model is trained on structured JSON captions, providing unprecedented control over composition, style, lighting, and color palettes. It also allows for precise object and text placement using bounding box controls
.Post-Generation Editing Suite
Ideogram 4.0 introduces a comprehensive suite of tools to revise and refine images after generation without starting from scratch:
Layerize Text: Turns generated typography into editable layers so you can restyle, reposition, and refine the text afterward.
Prompt Edit: Allows users to describe changes in plain language to revise specific parts of the image.
Extend & Reframe: Grows the composition beyond its original crop or shifts it for different aspect ratios while maintaining the original focal point.
Magic Fill: Adds new objects or replaces missing areas to seamlessly match the surrounding scene.
Other Tools: The platform also includes native background removal, upscaling for sharper delivery, and remixing to explore new creative directions.
🤗🙏🏼 Thanks to Ideogram AI and Kijai
Original repo — GitHub
Description
NEW v1.1.1 from 22.06: added SageAttention function and links, feel 2x acceleration!
Added Image reference for LLM function - get JSON prompt from any image and use it in Prompt Builder
Added shift parameter for quality extra management
Fixed some bugs
Updated tips
FAQ
Comments (9)
This works well for me except that every time I run it looks like the prompt is reset. Not huge deal because I can drag a previous png file into comfy, but it is a pain point.
@alfrodull954 Hi! What do you mean the prompt is reset? Do you use LLM enhancer? In my workflow you should use Manual Random Seed in the node. In this case each next generation will be different. If you want to save seed and for example make some changes like set other cfg or shift you should just push standart Run button in Comfy. In this case seed will remain the same but you will see the difference according to your settings
I just mean that the workflow by default is the "Ideogram is friendly then ever" picture. I change the prompt, get rid of the boxes and replace them with my own, and as soon as I hit the button to generate my image, everything resets back to the "Ideogram is friendly then ever" prompt and boxes. So if I want to tweak my prompt, I have to drag the generated image into comfy so it loads that image's workflow. Then I tweak it to generate a new image. But as soon as I hit generate, everything resets back to "Ideogram is friendly then ever".
Got it. In Step 1, there's a green prompt box just labelled "Prompt." That's where the reset is coming from. Once I deleted that text and saved the new workflow it stopped resetting.
@alfrodull954 you should just choose mode "When empty" in Prompt Builder node after it gets json prompt. After that you can manage everything in Ideogram 4 Editor window and changes will be saved
Nice workflow. Adding images and enabling reference image for LLM did not work for me. the prompt preview at the lower left (json prompt output) got a red outline and the prompt creation was skipped it seems. I've used the gemma model that's listed in the workflow. maybe something needs to be added for it to have vision? Edit: It worked the next time I opened the wf in comfy.
Also another question: Its probably recommended to deactivate llm prompt enhancer once you have your prompt and are just fine tuning it in the KJ ideogram node? :)
@Mech4nimaL Hi! Thanks! You can keep LLM enhancer turned on. If you turn on Prompt Builder and set mode When Empty, LLM will no longer affect the generation. In this case you can manage everything in your composition. When you need new LLM prompt you should return the mode Always in Prompt Builder. So it can get new JSON prompt
@RusselX but it seemed when llm enhancer is on it always runs even with the same seed, so it will take longer and you dont need it, when the prompt is already inside KJ prompt builder?
also sage attention threw an error in the 1.1 workflow, I will have a further look into it, because normally (in ltx etc) I'm using sage attention on auto without problems.
@Mech4nimaL So in that case you can turn off LLM enhancer if Prompt Builder already gets json prompt. And turn on it when you need new prompt.
Ideogram 4 works only with new SageAttention 2.2.0 post 5. Links and instructions how to install are in note on step 3