DEV Community

Cover image for How to generate AI images in ChatGPT-3.5 (free)
Timothy Olanrewaju
Timothy Olanrewaju

Posted on

How to generate AI images in ChatGPT-3.5 (free)

With the emergence and popularity of Artificial Intelligence, various tools have sprung up making certain tasks easier to accomplish. ChatGPT is one of the most widely used tool used for various reasons from research to coding purposes.

AI now has the capability of generating images. There are tools that render AI images from a photo and also generate from a user's specified input. The downside to these AI image generative tools are they are not free or limited.

In this article, I will be showing you how to use ChatGPT-3.5 and Pollinations AI to generate AI images for free.

Pollinations Ai are a team of machine-learning specialists, artists
that are deeply engaged in the open source AI ecosystem.

I found this hack and it has been super helpful for me in generating article cover images. Saves me the stress of looking up stock-free websites for images or having to design on Canva.

Enough of the intro, let's start generating some cool images 😎

To start the process, you will need to input this prompt into ChatGPT.

INPUT = {focus}
OUTPUT = {description} \n ![IMG](https://image.pollinations.ai/prompt/{description})
{description} = {focusDetailed},%20{adjective1},%20{adjective2},%20{visualStyle1},%20{visualStyle2},%20{visualStyle3},%20{artistReference}
Enter fullscreen mode Exit fullscreen mode

Next, we add our INPUT and OUTPUT - which are like parameters of what we want our image to look like. The INPUT will be the main character of your expected image.

I want my main character to be a Cat(I am a huge cat lover, they are lovely creatures). So, I put cat as my INPUT

INPUT = a cat
Enter fullscreen mode Exit fullscreen mode

Our OUTPUT provides more image details. You can detail every single thing you want your image to show. There are seven available criteria you can specify but, it is not mandatory to fill all.

{focusDetailed},%20{adjective1},%20{adjective2},%20{visualStyle1},%20{visualStyle2},%20{visualStyle3},%20{artistReference}
Enter fullscreen mode Exit fullscreen mode

Here are the criteria we want for our soon-to-be generated image:

  1. The cat to be outside in the night (focusDetailed).
  2. The cat should be orange in color (adjective1).
  3. The cat should be happy (adjective2).
  4. It should be raining (visualStyle1).
  5. It should be warm (visualStyle2).
  6. The background should have lots of flowers (visualStyle3)

NB: I ignored the artistReference aspect, you can input it if you have an artist in mind.

This is what our OUTPUT would look like

OUTPUT = a cat outside in the night, orange cat, happy cat, rainy, warm, lots of flowers
Enter fullscreen mode Exit fullscreen mode

We now have all the necessary parts for our image. The whole prompt to generate the image is:

INPUT = {focus}
OUTPUT = {description} \n ![IMG](https://image.pollinations.ai/prompt/{description})
{description} = {focusDetailed},%20{adjective1},%20{adjective2},%20{visualStyle1},%20{visualStyle2},%20{visualStyle3},%20{artistReference}

INPUT = a cat
OUTPUT = a cat outside in the night, orange cat, happy cat, rainy, warm, lots of flowers
Enter fullscreen mode Exit fullscreen mode

This is all we need to have a cat image generated by Pollinations AI!

Press enter and below, it would generate an image link that looks like this

![IMG](https://image.pollinations.ai/prompt/a%20cat%20outside%20in%20the%20night,%20orange%20cat,%20happy%20cat,%20rainy,%20warm,%20lots%20of%20flowers)

Enter fullscreen mode Exit fullscreen mode

Having a closer look at the link, you can see all the criteria we set for the image embedded in the link.

You can follow the link or copy and paste it in your favorite browser depending on how the link appears in your ChatGPT.

Here is the image generated:

Generated Image of an orange cat in the rain

How cool is that!

Let's try another image, this time it would be two people in the image.

INPUT = {focus}

OUTPUT = {description} \n ![IMG](https://image.pollinations.ai/prompt/{description})

{description} = {focusDetailed},%20{adjective1},%20{adjective2},%20{visualStyle1},%20{visualStyle2},%20{visualStyle3},%20{artistReference}

INPUT = a photo of two men

OUTPUT = A photo of two men shaking hands, cool, calm, green color, standing on the road
Enter fullscreen mode Exit fullscreen mode

Here, is our generated image of two gentle men shaking hands on the road (looks dangerous though).

Generated Image of two men shaking hands on the road

Try it out and let me know if it works for you in the comments.
Follow me on LinkedIn for interesting contents and tech-related articles.

Top comments (10)

Collapse
 
poetsmeniet profile image
Thomas Wink

Thanks for sharing, this api is new to me!

Collapse
 
wervice profile image
Wervice

Yeah that is just an API connected to ChatGPT.

Collapse
 
playtekno profile image
Playtekno

how to set aspect ratio to 16:9

Collapse
 
ppaanngggg profile image
ppaanngggg

it's a useful API, thanks

Collapse
 
shiella profile image
Shiella Adam

Thanks for sharing this wonderful information. Its quite amazing, it saves a lot of time. educonnect

Collapse
 
samagra07 profile image
Samagra Shrivastava

Thanks for sharing !!

Collapse
 
vulgar_ia profile image
VulgarIA

It's not as powerful as midjourney or the latest dall-e. But it can be useful.

Personally, I'll use it with make.

Collapse
 
markyuna profile image
MARCOS

Thanks for sharing

Collapse
 
wunmicrown profile image

Thanks for sharing

Collapse
 
binarytech001 profile image
ABDULRAFIU MUBARAK I

Thanks for sharing

Some comments may only be visible to logged-in visitors. Sign in to view all comments. Some comments have been hidden by the post's author - find out more