Why is it hard to create cards manually (one-by-one) and what can be the solution for this?

In board game design there are some time consuming processes like the creation of the cards. Did you now that this process can be automated? If you have a table or CSV file with the cards data (eg. Title, serial number, neccessary icons, images) and some basic images, then ImageMagick and Python can help you to create the cards. Imagine if you have 100 or more cards with different text on it and you decide to change the font or the background color. Instead of doing this manually for all 100 cards you can simply edit and run a Python script to update your cards.

In this article we will use the powerful and free ImageMagick tool which has a lot of image editing functions. Just check this site for examples, but avoid to deep dive into the documentation at this moment.

First, we write a Python script with the help of chatGPT.

Then we run the script and check the result.

Finally I give you a brief explanation of the most important parameters to understand which ones you need to edit to reach the expected result.

I will guide you through the whole process step-by-step.

It is good if you have some basic programming experience, but it is not a requirement to follow this tutorial.

If you are not familiar with programming or scripting or you just search for a more user friendly method, then jump to the end of this guide where I give you alternate ways to create the cards.

Tools we will use

• ImageMagick,
• Python,
• chatGPT,
• Notepad++ (or other IDE)
• Google sheets or a CSV file

Steps before building and running the script

Installing Python

Go to python.org and download the latest version. Follow the steps of the installation wizard. Before selecting Install Now turn on “Add python.exe to PATH”.
This option gives you the ability to run Python scripts in any folder on your computer.

Make sure to check “Add python.exe to PATH”

Installing ImageMagick

Go to the ImageMagick website and select the appropriate version of the software.
I use Windows, so I downloaded and installed the Windows version.

Make sure to check the “Add application directory to your system path” option

Installing a text editor, like Notepad++

I recommend to use Notepad++, because it has many useful features, like syntax highlighting. Go to the Notepad++ download page and install it.

Remark: You can use any other text editor or IDE. The simplest text editor in Windows is Notepad, but it don’t have the syntax highlighting feature. Another popular and powerful IDE (Integrated Development Environment) is Visual Studio Code (shortly: vscode).

Creating the project folder

Use any file management tool (or command line) to create a project folder with any name you like. I created the create_cards folder for this example project.

Create a subfolder with the name images in the project folder. We will put our basic images into this folder.

Create the icons subfolder in the project folder and place the text bubble icon into this folder.

After the project folder and the cards subfolder are created move on to the next step.

This is the folder structure after all necessary folders are created:

Remark: The output folder will be created by the Python script.

Create or download the starting images

We need to put the starting card images into the cards folder and the message.png icon into the icons folder.

You can download them in one .zip file or one-by-one below.

message64.png:

message64.png (white background at the text bubble)

Generating the first script

I use chatGPT to generate the Python code.

This is the prompt to generate the script (feel free to use other prompt to customize the script for your current project):

Help creating a Python script which uses ImageMagick to create cards for a fantasy themed card game. The card with is 45mm and the height is 68mm. The resolution is 150dpi. Use a CSV file as a data source for the cards. The CSV file contains the title, serial number, icon and base image. Place the title in the upper center part of the card in yellowish gold color on darkgreen background. Place the serial number in the upper left corner with black text and use an orange circle with black border as background. If the third column (icon) in the CSV file is True, then place the message64.png icon from the icons folder into the lower left corner on the card. Place the base image in the center of the card. Use darkblue background for the card. The base images are in the images folder. Save the cards into the cards_output folder. Be aware of how ImageMagick works: “Image Composition is the process of merging two (and only two) images in a variety of ways.”

So my prompt contains the following parameters:

• Using a CSV file
• Card size in millimeters
• Print resolution in DPI (I use 150dpi for faster process and testing)
• Using some images without any text and number (center align)
• Card title (top align)
• Card serial number in an orange circle
• Message icon – if needed on a specific card

Tip: Be as concrete as you can when you formulate your prompt. Then the result will we more close to your wishes.

Script and CSV file what I got

Here you can see the script (double click on the code to select all):

import csv
import os
import subprocess
import tempfile
import shutil

# Constants
WIDTH_MM = 45
HEIGHT_MM = 68
DPI = 150
WIDTH_PX = int(WIDTH_MM / 25.4 * DPI)
HEIGHT_PX = int(HEIGHT_MM / 25.4 * DPI)

OUTPUT_DIR = "cards_output"
IMAGES_DIR = "images"
ICONS_DIR = "icons"

os.makedirs(OUTPUT_DIR, exist_ok=True)

def run_magick(args):
    subprocess.run(["magick"] + args, check=True)

def generate_temp_image(filename, args):
    temp_path = os.path.join(tempfile.gettempdir(), filename)
    run_magick(args + [temp_path])
    return temp_path

def create_card(title, serial, icon, base_image_filename):
    base_image_path = os.path.join(IMAGES_DIR, base_image_filename)
    icon_path = os.path.join(ICONS_DIR, "message64.png")
    output_path = os.path.join(OUTPUT_DIR, f"{serial}.png")

    temp_files = []

    try:
        # 1. Base background (dark blue)
        base_bg = generate_temp_image(f"{serial}_bg.png", [
            "-size", f"{WIDTH_PX}x{HEIGHT_PX}", "xc:#001a4d"
        ])
        temp_files.append(base_bg)

        # 2. Title bar with text
        title_img = generate_temp_image(f"{serial}_title.png", [
            "-size", f"{WIDTH_PX}x40", "xc:#004d00",
            "-gravity", "center", "-pointsize", "20", "-fill", "#FFD700",
            "-annotate", "+0+5", title
        ])
        temp_files.append(title_img)

        composed1 = generate_temp_image(f"{serial}_step1.png", [
            base_bg, title_img, "-gravity", "north", "-composite"
        ])
        temp_files.append(composed1)

        # 3. Serial number circle
        serial_circle = generate_temp_image(f"{serial}_serial.png", [
            "-size", "100x100", "xc:none",
            "-fill", "orange", "-stroke", "black", "-strokewidth", "4",
            "-draw", "circle 20,20 50,0",
            "-gravity", "northwest", "-extent", f"{WIDTH_PX}x{HEIGHT_PX}",
            "-gravity", "northwest", "-pointsize", "20", "-fill", "black",
            "-annotate", "+25+30", serial
        ])
        temp_files.append(serial_circle)

        composed2 = generate_temp_image(f"{serial}_step2.png", [
            composed1, serial_circle, "-gravity", "northwest", "-composite"
        ])
        temp_files.append(composed2)

        # 4. Center image (base image)
        resized_base = generate_temp_image(f"{serial}_baseimg.png", [
            base_image_path, "-resize", f"{int(WIDTH_PX*0.8)}x{int(HEIGHT_PX*0.5)}"
        ])
        temp_files.append(resized_base)

        centered_base = generate_temp_image(f"{serial}_basecenter.png", [
            "-size", f"{WIDTH_PX}x{HEIGHT_PX}", "xc:none",
            resized_base, "-gravity", "center", "-composite"
        ])
        temp_files.append(centered_base)

        composed3 = generate_temp_image(f"{serial}_step3.png", [
            composed2, centered_base, "-gravity", "center", "-composite"
        ])
        temp_files.append(composed3)

        # 5. Optional icon in lower left
        if icon.lower() == "true":
            resized_icon = generate_temp_image(f"{serial}_icon.png", [
                icon_path, "-resize", "40x40"
            ])
            temp_files.append(resized_icon)

            icon_layer = generate_temp_image(f"{serial}_iconlayer.png", [
                "-size", f"{WIDTH_PX}x{HEIGHT_PX}", "xc:none",
                resized_icon, "-gravity", "southwest", "-geometry", "+10+10", "-composite"
            ])
            temp_files.append(icon_layer)

            composed4 = generate_temp_image(f"{serial}_step4.png", [
                composed3, icon_layer, "-gravity", "southwest", "-composite"
            ])
            temp_files.append(composed4)

            shutil.copy(composed4, output_path)
        else:
            shutil.copy(composed3, output_path)

        print(f"Card created: {output_path}")

    finally:
        # Clean up temp files
        for file in temp_files:
            try:
                os.remove(file)
            except FileNotFoundError:
                pass

# Read CSV
with open("cards.csv", newline='', encoding='utf-8') as csvfile:
    reader = csv.DictReader(csvfile)
    for row in reader:
        create_card(
            title=row["title"],
            serial=row["serial"],
            icon=row["icon"],
            base_image_filename=row["base_image"]
        )

Here is the csv file:

title,serial,icon,base_image
Spartan Soldier,01,True,spartan.png
Titan Warrior,02,False,titan.png
Elf Wizard,03,True,elf.png
Dark Goblin,04,False,goblin.png

Saving and running the script

chatGPT generated the Python script and a sample .csv file (see above).

Create a new Python file in your project folder, copy-paste the script and save it.
Remark: you can create a new file in Notepad++, paste the Python script and save as create_cards.py file.

Create a new CSV file and copy-paste the sample CSV file content from chatGPT or the code above. Save the file as cards.csv.

Run the script simply by double clicking on the filename.

If there are no errors, then the script runs and saves the card images into the cards_output folder. Check the folder for the results.

If there are errors in the Python script, then you should fix them to run the image export. See the Troubleshooting chapter to learn how to fix errors.

Tip: if you would like to see what Python writes on the screen, then do one of the following:

1. run the script in command line. To do this open the command line by hitting the Windows button and type cmd, then hit ENTER. Navigate to the folder of the script (use cd foldername) and type the Python program name to run it.
2. Insert the following line of code after the last line in the script in a new line and save the file:
   input(“Press ENTER to exit the program!”)

Editing the script

Right click on the .py file and select “Edit with NotePad++”. Be sure to save the file after editing.

Troubleshooting the script

Python prints error messages on the screen if the script contains errors. I help you fix these errors in this chapter.

“No such file or directory” error

magick: unable to open image ‘images\spartan.png’: No such file or directory @ error/blob.c/OpenBlob/3596.

Solution:

• copy the images to the correct folder
• rename the images if the names are incorrect

Fine-tuning the results

Fixing object size, transparency settings and alignment

I made several fixing steps.
To fix transparency I replaced row 60 with this line:

        "-gravity", "northwest", "-background", "none", "-extent", f"{WIDTH_PX}x{HEIGHT_PX}",

These are the thumbnails of the test images:

More tweaking needed:

• decrease the radius of the orange circle
• decrease the border of it
• fix the background of the message64.png icon (you can download the fixed version above)

These are the thumbnails of the refined images:

You can download the refined Python script from this link.

Changing text color

I checked the exported images and I see, that the text is too bright so I change the code:

Sorry, this chapter is just a draft. Update coming soon…

Different aspect ratio

If you would like to create 1:1 aspect ratio images, then simply write the same number for width and height (row 8 and 9 in the Python script, CARD_WIDTH_MM and CARD_HEIGHT_MM).

Using Google fonts

You can easily change the font by replacing the approriate line:

Sorry, this chapter is just a draft. Update coming soon…

I selected the Fancy Feeling on fonts.google.com and browsed through the fonts. My final decision was the Birthstone font.

Always read the license of the fonts to learn about the usage of them.

Creating exact number of some type of cards

Sometimes we need to create almost the same cards but with different text, icons and numbers.

Imagine that we would like to create a deck of 30 cards. 10 cards with a dragon, 10 cards with a knight, 5 cards with dwarfs, 5 cards with a sword.

Then instead of using different images for every card, we shoud include another column in the table or in the CSV file, which indicates which image we need on which card.

Sorry, this content is just a draft. Update coming soon…

Prepare the cards for print

Now we have separate cards with nice text and icons. The next step is to print them. Use the following script to align more then one card on a page with. I use A4 paper size, but you can use any other standard. A4 paper dimensions: 210x297mm.

Sorry, this chapter is just a draft. Update coming soon…

Alternative ways to create cards for board games

If you don’t like scripting and you would like to do this job with a more user friendly way, than you can use one of this tools:

• Nandeck
• Dextrous
• Forge of Cards
• Tabletop Creator

Dextrous is very user friendly, you see the end result immediately and it can use .CSV files as data source for card data. The free version can be enough for most of the use cases.

The list is not complete and I may update it later.

Final words

I hope you enjoyed this guide and found it useful. Happy card designing! If you have any questions feel free to ask them in the chat below!

Thanks for reading this long article! 🙂

The post Creating cards for board game with ImageMagick and Python appeared first on AI Art Blog.

So from now on, image generation is not done by Dall-E by default in ChatGPT, but by the GPT-4o model. This development has also been integrated into the Sora video generator. The descriptions and demonstration images are quite promising, so I thought I would try what this tool can do. Update: it is no longer available in the free version, so I have not tested the service yet. I will only highlight a few official images.

According to OpenAI, a good image generator should not only be able to create spectacular photorealistic, fantasy, and surreal images, but also be useful and functional.

When creating the new image generation model, attention was paid to ensuring that the texts appearing in the generated images were accurate. The model not only monitors the correctness of the text, but we can also specify the exact arrangement of the text, so we can write poems or create beautifully illustrated recipes or infographics.

The model also supports differentiated imaging, so after image generation, we can easily modify the existing image by entering a short prompt.

Furthermore, this model can easily create comics and even user interfaces (UI) for games with full visuals and consistent icons.

This latter feature could be a game changer, as it will now be possible to create UIs and infographics with precise text much faster and more efficiently.

Sample images

Unfortunately, access to the new image generator has now been disabled in the Free version, so I’ll only show some of the official images:

Example #1 – Infographic

Image credit: OpenAI

Correct captions and convincing depiction.

Example #2 – Reflections and text

Image credit: OpenAI

The reflections look quite accurate. Many AI image generators have trouble representing reflections properly.

Example #3 – UI design

Image credit: OpenAI

We can create computer game visual designs even with consistent icons.

Example #4 – Transparent background

Credit: OpenAI

https://openai.com/index/introducing-4o-image-generation/?video=1069289433

Example #5 – Illustration for a science experiment

Image credit: OpenAI

Availability

The gpt-4o image generator is now available on Plus, Pro, and Team subscriptions. Due to overwhelming demand, the ability to generate images with the new model was disabled in the free version to prevent server overload.

The service will soon be available in Enterprise and Edu versions.

Performance

This model is a bit slower than Dall-E, as it interprets and processes our prompt more precisely. It can take more than 1 minute to render an image.

Advanced prompting

Since there is no UI element to set parameters (such as aspect ratio, colors used, image style) during image generation, all of this can be specified simply in text form. You can even specify the hexadecimal code of specific colors in the prompt. If you want to create a photorealistic image, you can specify the type of camera or lens or even the style of the photo (example: polaroid photo).

Limitations

The model does not always work flawlessly. There may be problems with rendering multilingual text, hallucinations may occur. Longer, small texts may be incorrectly rendered. There may be problems with content authenticity if we want to display more than 10-20 small icons in an image. We can see examples of this in the official article.

Summary

Based on the sample images on the official website, I can say that the gpt-4o image generator can be a very useful tool for creating illustrations, infographics, UI elements and icons.

Of course, we have to be careful if we want to use the images for commercial purposes. If we have the image made in the style of a specific artist, it can lead to legal problems. This applies to any AI image generator.

The post GPT 4o image generation is finally here appeared first on AI Art Blog.

Blog Updates and New Initiatives

This year, the blog expanded in exciting directions. A few highlights include:

• AI Timeline Page: I created a dedicated timeline to track the progression of AI milestones. This page serves as a historical reference and showcases the rapid evolution of generative AI.
• AI Art Gallery Series: 2024 marked the launch of the AI Art Gallery series. These posts celebrate the creativity unlocked through generative AI tools, featuring thematic artwork crafted by various platforms.
• GenAI Power Consumption: A recurring theme this year has been the environmental impact of AI. I wrote an article about the importance of the energy demand of generative AI.

Breakthroughs in AI Text Generation

Generative AI continued to dominate headlines with significant updates from major players:

• GPT-4o by OpenAI: OpenAI introduced GPT-4o, an optimized and more efficient version of its large language model, improving performance without compromising quality.
• OpenAI o1: This release focused on better reasoning, solving more complex tasks and problems, but “thinks” more, than the lightweight models. Read my article about o1.
• Google Gemini: Google’s new model, Gemini, combined cutting-edge text and image processing capabilities, setting a new benchmark for AI integration.
• Claude 3.5: Anthropic’s Claude 3.5 Sonnet brought enhancements in conversational AI, it has deep analytical capabilities, coding skills, creative writing and so on.

Advancements in AI Image Generation

This year also saw transformative updates in AI-powered imagery:

• Flux AI: A new player on the scene, Flux AI, introduced dynamic and realistic image-generation capabilities, drawing attention with its unique rendering styles.
• Fooocus AI: Fooocus AI emphasized precision and control, offering users the ability to refine outputs like never before.
• Stable Diffusion 3.5: The release of Stable Diffusion 3.5 continued to push boundaries, with improved image clarity, speed, and customization options.
• Playground AI’s Shift: Playground AI redefined its approach, transitioning from pure creation to a design-focused model. This pivot reflected broader trends in how AI tools are evolving to meet user needs.

The Rise of AI Video Tools

The generative video space made waves with notable launches:

• OpenAI’s Sora: This innovative video tool expanded content creation possibilities. With support in various countries (check the list here), Sora is democratizing video generation.
• Other Generators: Several other platforms also emerged, for example Runway or the Flux AI Video Generator.

Innovations in AI Music

Music creation saw significant strides, making generative AI an indispensable tool for musicians and producers:

• Suno Updates: The platform introduced a range of new features, culminating in the release of Suno v4, a groundbreaking update that expanded creative possibilities for music generation.
• Udio’s Evolution: Udio added several new features, focusing on enhancing user experience and output quality.

Looking Ahead to 2025

As we move into 2025, the possibilities seem boundless. Here’s what I’m keeping an eye on:

• Sustainability in AI: Expect more emphasis on eco-friendly AI development, as power consumption remains a hot topic.
• Integration and Accessibility: The trend towards lightweight, integrated AI tools will likely continue, breaking down barriers to adoption.
• Creative Frontiers: The convergence of AI with art, music, and video will further blur the lines between human and machine creativity.

2024 has been a remarkable year. The generative AI trends showed in 2024 that many new innovations was born. I’m curious to see what 2025 has in store. Thank you for following along on this journey. Let’s continue exploring, questioning, and celebrating the incredible advancements of genAI!

I wish you a Happy and Creative New Year!

The post Generative AI Trends in 2024 appeared first on AI Art Blog.

Who is this service available to?

Search is available to ChatGPT Plus and Team subscribers, so free users cannot use it for now. On their website, OpenAI promises to make this feature available to all users in the coming months:

“We’ll roll out to all Free users over the coming months.”

What are the advantages of ChatGPT search?

The advantage of ChatGPT search is that it integrates the capabilities of AI-based answering and traditional web search, often resulting in more accurate and targeted results. Instead of just providing a list of results like Google, ChatGPT is able to contextualize information, summarize it and directly answer the user’s questions. This can be useful if the user has specific questions and does not want to waste time going through the search results one by one.

Example: we want to travel to a sandy beach in Europe. In addition to specific offers, you can further refine your search and search for local food specialties and the expected weather.
In addition to text results, ChatGPT search can also display images, tables, graphs, maps and other formatted data.

Can ChatGPT replace Google search?

While OpenAI’s new feature looks useful, it’s unlikely to replace Google search any time soon. Google provides a much wider range of search results and a more detailed list of results, which many users like. ChatGPT search, on the other hand, is ideal for quick, accurate answers, especially when the user asks a specific question. It can be more of a complement to Google than a replacement, especially for users who need quick answers without digging deeper into search results lists.

Of course, it is impossible to predict how popular this service will be in the future, but the fact is that an interesting innovation has been added to ChatGPT.

The post Could ChatGPT search be Google’s biggest rival? appeared first on AI Art Blog.

Who Can Try the o1 Model?

At the moment, the o1 model is available to ChatGPT Plus and Team users, with Enterprise and Educational users getting access soon after. If you’re subscribed to ChatGPT Plus, you can already experiment with two versions of the model: o1-preview and o1-mini. The first one is the flagship model focused on solving particularly challenging tasks, while o1-mini is a more affordable and faster alternative, especially useful for coding-related tasks.

There are some limitations, though, such as message caps:

• 30 per week for o1-preview and
• 50 per week for o1-mini

but these are likely to change as OpenAI refines the model further.

When Will It Be Available for Everyone?

For now, the public can’t try o1 unless they’re paying for the Plus or Enterprise tiers. However, OpenAI has mentioned plans to eventually roll out access to ChatGPT Free users, especially the o1-mini version, which is both faster and cheaper. This might happen in the coming months, but there’s no fixed date yet. It’s something worth keeping an eye on it, if you would like to try it, but don’t want to pay for subscription just yet.

Videos Worth Watching

On OpenAI’s official o1 page, there are some videos that demonstrate the power and flexibility of this new model. Here are three that caught my attention:

HTML Snake Game
In this video, o1 helps create a classic snake game in HTML. It shows how the model can handle multi-step coding tasks efficiently. This is particularly interesting because it demonstrates o1’s ability to generate and debug code for developers quickly. Even if you’re not into game development, it’s worth watching to see how the model works through the logic step by step.

Solving a Complex Logic Puzzle
This video showcases o1’s reasoning abilities by tackling a complex logic puzzle. What’s fascinating here is how o1 constructs a detailed “chain of thought” before delivering its final answer. It gives you a peek into how the model breaks down intricate problems, something that GPT-4o wasn’t as strong in.

Quantum Physics Challenge
Ever wondered how an AI would handle quantum physics? In this video, o1 goes through the paces of generating complex mathematical formulas needed for quantum optics. This is where o1 really shines—its ability to deal with the sort of difficult, niche problems that you wouldn’t normally throw at an AI. Watching it process these advanced concepts makes it clear why OpenAI is touting this as a breakthrough model for science and research.

I’m curious about how this new model handles basic mathematical tasks, like counting letters in a word. The previous models were not too good at calculations, but if I ask GPT-4o to write a program for this task, it generates a well-functioning script.
The o1 seems to be able to do simple math operations that the GPT-4o couldn’t.

Why Try Out OpenAI o1? Is It Better Than GPT-4o?

So, what makes o1 worth trying? For one, the model is designed to spend a little more time thinking before it answers. This extra computational effort makes a huge difference in fields like math, physics, and even coding, where the ability to reason and understand the problem before generating an answer is crucial.
For example, o1 outperformed GPT-4o in solving math problems at the International Mathematics Olympiad, solving 83% of the problems compared to GPT-4o’s 13% success rate.

Moreover, o1 is highly specialized for tasks that require multi-step reasoning, like generating scientific hypotheses or debugging complex code. While GPT-4o is still useful for broader, more generalized tasks, o1’s ability to “think” deeply makes it an excellent choice for more technical challenges.

Despite these advantages, there are limitations:

• As an early model, o1 lacks some features that make GPT-4o practical for everyday use, such as browsing the web or handling files and images. So, depending on your needs, GPT-4o might still be the better choice for now if you’re aiming to use these capabilities.
• The operational costs of using the o1 model are significantly higher than those of GPT-4o, with input costs three times greater and output costs four times higher. Additionally, processing speed can be slower for complex queries.

Summary

While I haven’t personally tested OpenAI o1 yet, the early demonstrations and use cases suggest it’s a game-changer for anyone working in fields that require complex problem-solving. Whether you’re a researcher tackling quantum mechanics or a developer debugging code, o1 seems to have the potential to be much more than just a language model—it’s a tool for reasoning.

Right now, only those on ChatGPT Plus or higher tiers can test it out, but with plans to make it accessible to Free users, it’s something I’d recommend keeping on your radar.

The post OpenAI o1: Better Reasoning in complex tasks appeared first on AI Art Blog.

There are a lot of changes, new functions in these free AI image generators. I can mention some new and powerful AI image tools which are released this year.

There are some changes, for example OpenAI enabled to use DALL-E 3 free, but with limitations.

Playground AI made a huge update in 2024: v3 beta is available.

Black Fores Labs introduced Flux image generator.

Midjourney is unfortunately not free due to the high demand and popularity, you must pay for the subscription.

DALL-E 3

https://openai.com/index/dall-e-3/

OpenAI was among the first to create an AI image generation tool. The free version of DALL-E 3 became available a few weeks ago in ChatGPT. You can learn from this article how to create images with different aspect ratios.

There is an another way to generate images with DALL-E 3 for free: it is embedded into Microsoft Copilot as well.

General information:

• You can create 2 images per day with ChatGPT

Benefits of the free version:

• easy to use
• ChatGPT can enhace your prompt to make it more detailed and accurate

Limitations of the free version:

• changing the parameters is not easy

Playground AI

https://playgroundai.com/

Playground AI made a huge update in this year. They released v3 beta and implemented the so called Design function.

If you navigate to the website you can first select a project type (for example Logo, Seamless Patterns or Social Media Post) and after that you can choose from a large list of templates. Many templates are free, but some are available only with Pro subscription.

If you selected the template, than you can change the image and change the style.

I’m plan to write an article about this brand new design tool, so check back later.

You can generate images with the original create tool of course.

General features:

• User gallery (Community Feed)
• Possibility to remix generated images by others
• Keyword search
• Canvas editor
• Designer
• Ability to change image aspect ratio
• Various filters

Limitations of the free version:

• 10 images in 24 hours
• 30 prompt expansions / month
• 5 creative upscales / month
• 30 subtle upscale / month
• Maximum resolution of generated images: 1024 x 1024 pixels in case of 1:1 ratio
• Slower image generation, the Pro
• Other image quality restrictions
• Not all design templates are available for free

Leonardo Ai

https://leonardo.ai

Leonardo Ai is even better then in 2023. In this year some new innovative features was added to this tool, like the Phoenix preset. In the early days of computer generated images it was a big problem to render text on the image, but Leonardo Ai solved this problem with Phoenix.

General features:

• User gallery (Recent Creations)
• Possibility to remix generated images by other users
• Canvas editor
• Realtime Canvas (originally called live canvas, see my article and youtube video)

This application has additional extra features such as:

• texture generation for 3D models
• video generation

Limitations of the free version:

• 150 fast tokens, resets once per day
• no private generations
• Images can be up to 1024×1024 pixels (in case of square ratio)

Flux

https://blackforestlabs.ai

There is a new AI art generator tool on the market created by Black Forest Labs. The rumor is that this is a revolutionary AI image generator because it can create photorealistic images.

I heard that the people worked on Stable Diffusion made Flux as well.

If you scroll down on the Flux website then you can see that many images seems to be photorealistic in thumbnail size and looks even as a real photo if you enlarge it. So Flux is very convicing, I will try it soon.

Interesting is that it’s open source, so everyone can download it for free and experiment with it.

Flux has three different versions:

• pro – for professionals – this is unfortunately not free
• dev – still very quality model, but only for non-commercial usage
• schnell – the fastest model, for personal use, you can install and run it on your own computer and generate images in a couple of seconds depending on your computer speed

Summary

The free AI image generators mentioned above providing a good possibility to try AI-based image generation. If you’re on a limited budget or just curious to see what these apps can do, it’s worth trying at least one of them out. I personally recommend Playground AI and Leonardo Ai because of the wide variety of customisation possibilities even in the free version.

I would like to see your creations! Feel free to try the apps above and share your results!

The post Best free AI image generators in 2024 appeared first on AI Art Blog.

With this step, the company has achieved that you can also test this tool (if you are registered on chat.openai.com) and generate two images per day with the AI. DALL-E 3 is available from September for ChatGPT Plus subscribers.

I also tried this quickly and in this article I will present the results.

The tests

I will also share the pictures and the prompt with you. The pictures are original size and I have not done any post-processing on them.

Test #1

In the first test, I wanted to examine both how well DALL-E understands the prompt I wrote and what texts it can produce on the car.

I gave this prompt: “Draw a racing car in side view, racing on the race track, its wheels and the background are blurred by the high speed, the car’s side is supported by the inscriptions “chatGPT”, “Dall-E 3”, sunset, golden hour, unreal engine, hyper realistic“

The result is of good quality, it brings the atmosphere I imagined. The texts are not perfect, we can correct them afterwards with an image editing software.
(It is possible that the caption would already be accurate after multiple generations.)

Test #2

The second image generation option was inspired by a book I read about Norse mythology. My idea was to represent the tree of life, Yggdrasil.
Another interesting thing is that I did not give a detailed prompt, but only a few words.

The short prompt: “Draw an oil painting of the Yggdrasil from the Northern mythology.“

Which chatGPT automatically replaced with a long, detailed description with four paragraphs (I inserted the first and last paragraph only):

“Imagine an oil painting of Yggdrasil, the immense and mystical ash tree that connects the Nine Worlds in Norse mythology. The tree stands tall and grand, its sprawling branches reaching up to the heavens, filled with lush green leaves that shimmer with an otherworldly glow. The trunk is thick and gnarled, with intricate patterns etched into the bark, symbolizing the ancient runes of wisdom...

The painting exudes a sense of ancient power, mystery, and the interconnectedness of all things in the Norse cosmos, with vibrant colors and rich textures that convey the majesty and sacredness of Yggdrasil.“

The picture turned out to be very atmospheric, although to me it looks more like a digital painting than an oil painting.

How to generate image with custom aspect ratio?

With a very easy trick you can customize the image aspect ratio, so you are able to create not only 1:1 images but also, for example header images (landscape format) or portrait format images.

The trick is to indicate in the prompt which aspect ratio you want to see in the final image.

I used this prompt to generate the drawing robot image with chatGPT:
Create a 3:2 ratio vector illustration for 12-14 year old children: a humanoid robot drawing an image with colored pencil on a piece of paper which is on a desk. The robot is lightblue, the colors of the room are warm, the pencil has an orange color.

Summary

OpenAI’s goal is certainly to get as many people as possible to try the DALL-E image generator and to subscribe to chatGPT Plus to create unlimited images.

I recommend to try chatGPT DALL-E 3 for free, but think carefully about which prompt you use. Because you only have two options per day.

The post ChatGPT made DALL-E 3 free for all users appeared first on AI Art Blog.

This short video shows how amazingly realistic works of art can be created today with the help of some artificial intelligence tools.
You can watch the video at the end of the article.

Fukoe used the following tools to create the music video:

• ChatGPT for lyrics
• Suno to the music
• DreamMachine, Gen3 alpha and Kling for video production
• Midjourney and Stable diffusion for images

The created content was edited with Adobe Photoshop and After Effects software in the post production process. This is the only part of the workflow where artificial intelligence solutions were not required.

The end result is a convincing, photorealistic and at the same time surreal work.

In conclusion, Arata Fukoe’s music video “AI Evolution” is a significant achievement in the field of AI-generated art. Using advanced generative artificial intelligence tools, Fukoe created a music video that pushes the boundaries of traditional artistic expression. As artificial intelligence technology continues to evolve, it is likely to play an increasingly significant role in shaping the future of creative creation.

The post AI music video that was made almost entirely with generative AI appeared first on AI Art Blog.

In this article, we’ll take a look at what OpenAI’s latest GPT-4o model can do. If you’re curious about its capabilities, be sure to read on.

On Monday, OpenAI announced its newest, fantastic model named GPT-4o. They practically wrote an article and there are plenty of videos on their YouTube channel in which they demonstrate its capabilities.

We will go through them in broad outline, so that by the end of the article you will have a comprehensive picture of what makes this new large language model (LLM) so groundbreaking.

It is noteworthy that GPT-4o will also be available for free.

News about GPT4-o

Its most notable ability is that you can practically have a conversation with it. The reaction time for answering has been reduced to a minimum with this AI model, and you can also interrupt his speech.

The previously available Voice service of ChatGPT needed 4-5 seconds for the answer, now that has completely changed.

Watch this example video to see what’s new.

The GPT4-o model

In the description, you can read that this model is actually an omni model. This means, that it can not only interpret your voice (Voice mode) and text, but can practically handle both.

The interpretation of images and videos is made possible by the Vision mode.

The Memory mode remembers our previous conversations, which ChatGPT can use later. With the Analyze mode, complex data can be analyzed quickly and efficiently. Examples: Excel spreadsheets or even a chart that you show as an image to chatGPT.

Practical examples

If you look at the introduction page of GPT-4o, you can see many examples of people using their phone’s camera with the ChatGPT app and asking ChatGPT what it sees in the camera, and it then answers them “verbally”.

So the same thing can now be done in real time with your phone’s camera. The big difference between the current GPT-4o and the previous Voice service is that a completely new model was actually trained here with voice, text, videos and images.

It is also worth watching the video where they show that one person speaks Italian, the other person speaks English. The phone interprets it in real time. This means that as soon as the Italian text is spoken, ChatGPT immediately translates it into English and vice versa.

You’ve may seen conversations, where politicians spoke in their own mother tongue and still understood each other. It was thanks to the dubbing service in the background.

In the future, this will probably be completely replaceable with such an AI solution.

Performance of GPT-4o

There is a comparison chart on the hello-gpt-4o page in the Model evaluations section, with which we compare the GPT-4o model through various benchmarks, for example with GPT4 turbo, or even with the paid Claude 3 Opus and Gemini Pro 1.5.

The speed difference according to the graph will become visible in practice. All models are already very fast and are therefore capable of responding in real time.

You can feel the difference between a free ChatGPT, i.e. GPT 3.5 and GPT 4, but the difference between the above big models is not really noticeable anymore.

Security and limitations

The good news is that they also put a lot of emphasis on security. It’s highly likely that you won’t be able to say obscene words with it and you won’t be able to ask ChatGPT for other silly things either.

This model is subjected to serious, in-depth testing, which means that a team tries to hack these systems and extract answers from them. The name of the testing method is the so-called red teaming.

The future

These services are constantly being improved. Maybe by the time you read this article, the entire interface will look completely different.

Thanks to the Memory mode, ChatGPT will also remember previous conversations. You can turn this off if you don’t want it to store information about you.

The Analyze mode is good for uploading documents and requesting all kinds of analytics from it. So there will be improvements here in the free GPT as well.

Summary

This new model contains huge innovations. It almost feels like talking to a live person. It can also be used to facilitate scientific tasks, such as the analysis of complex data, graphs, and images.

If this is also combined with image and video generation, or perhaps with the generation of sound and music, the limit to what we can create with it is the imagination.

Of course, he won’t cook lunch for you, but he can help you in creative and scientific tasks.

Thank you for reading this article!

Image by Franz Bachinger, Pixabay

The post OpenAIs GPT-4o model is free to use! appeared first on AI Art Blog.