Best AI Image Generators for Beginners
Trying to figure out how to start generating AI images can be overwhelming.
Really overwhelming.
There are so many tools to choose from.
Here are the top AI image generators for you to consider as you are starting out. Depending on what is important to you, one may be for you better vs. another.
Note.
All of these AI image generators are first and foremost ‘text-to-image’ generators. That is, you give the AI model a prompt, basically text, that describes what you want to generate, then the AI model generates you an image.
Before we get into it – really quickly:
1. Who am I to talk about AI tools?
I have generated close to 200,000 AI images. I evaluate and put AI tools through rigorous testing. This is what I do.
2. There is a lot of information out there.
Outdated. Biased. Inaccurate.
My goal is to provide you the most objective, accurate, and unbiased information possible, so that you can make the best decision for you on what AI tool to get started on.
Ok, let’s go.
Top AI Image Generators for Beginners
DALL·E 3
Easy for beginners with ChatGPT and Microsoft Copilot integration. Great for a variety of images, not for photorealism.
Freepik
Intuitive and easy to use with customization options. Great for photorealistic images, more practical than artistic.
Midjourney
Initially more complicated, but it offers exceptional image quality and versatility. Fantastic for creative exploration.
DALL·E 3
If you’re already using ChatGPT Plus (paid account) or Microsoft tools, DALL·E 3 provides an easy entry into AI image generation.
If you want to generate photorealistic images, DALL·E 3 may not be the best option for you.
DALL·E 3 via ChatGPT
SPECS
Input method: conversational text
Number of images generated per run: 1
Default image size: 1024×1024
Aspect ratio: square, portrait or landscape
Output format: webp
Access: via website, mobile app
Generation time: 15-25 sec.
Prompting knowledge: helpful
Features: image editing (inpainting) NEW
Restrictions: can’t use living artists names in prompts
PROS
Ease of use: Easy, especially for those familiar with ChatGPT interface
Prompt adherence: very good
Can generate text in an image
CONS
Struggles in generating photorealistic images
Limited customization and image editing options
Minimal creative control
No centralized image archive
PRICING – via ChatGPT
No free plan.
Paid Plan
ChatGPT Plus: $20/month
DALL·E 3 via MS Copilot & Designer
SPECS
Input method: conversational text
Number of images generated per run: 4
Default image size: 1024×1024
Aspect ratio: Free – square, Pro plan – only landscape (16:9)
Output format: JPG
Access: website, mobile app (iOS and Android)
Generation time: 15+ sec.
Prompting knowledge: helpful
Features: edit images in Designer, customize images, add to design templates, apply filters
PROS
Ease of use: some learning curve
Prompt adherence: very good
Can generate text in an image
CONS
Struggles in generating photorealistic images
Often buggy (e.g. image generation can get stuck)
PRICING – via Microsoft Copilot
Free Plan
15 daily boosts, after boosts run out can still generate images, just takes longer
Paid Plan
MS Copilot Pro: $20/month
Faster AI image generation with 100 boosts per day. Also includes priority access to ChatGPT-4.
Examples – using ChatGPT
My input:
Create a futuristic city skyline at night, with a large illuminated building and a river sparkling below
The prompt ChatGPT created:
A futuristic city skyline at night, dominated by a large, uniquely designed, illuminated building that stands out against the dark sky. The city is filled with sleek, modern skyscrapers of various heights and designs, featuring glowing windows and neon accents. A wide river flows through the city, reflecting the dazzling lights from the buildings and the vibrant neon signs, creating a sparkling effect on its surface. The night sky above is clear, allowing the city’s illumination to shine brightly. This setting combines elements of advanced technology and architecture, suggesting a world where innovation and aesthetic beauty coexist harmoniously.
My input:
Create a photo of a delicious cheeseburger.
The prompt ChatGPT created:
A close-up photo of a delicious cheeseburger on a wooden table. The cheeseburger includes a juicy beef patty, melted cheddar cheese, fresh lettuce, sliced tomatoes, pickles, and onions. It is served on a toasted sesame seed bun, with a slight glisten on the beef and cheese to emphasize its juiciness and freshness. The background features rustic wooden textures to enhance the warmth and appeal of the meal, creating a cozy and appetizing atmosphere.
a photograph of a blond woman wearing a knitted wool hat and scarf, cyberpunk style
The prompt ChatGPT created:
A photograph of a blond woman wearing a knitted wool hat and scarf, styled in a cyberpunk aesthetic. The woman is posed against an urban, neon-lit background, typical of cyberpunk visuals. She has sharp, futuristic makeup with glowing neon accents matching the colors of the city lights. Her attire, though knitted, incorporates metallic and holographic elements, blending traditional warmth with the cold, high-tech environment of a cyberpunk world. The atmosphere is charged with a mix of old and new, organic and synthetic, showcasing a stark contrast and harmony between the elements.
Examples – using Copilot/Designer
Futuristic city skyline at night, with a large illuminated building and a river sparkling below
A photo of a delicious cheeseburger
a photograph of a blond woman wearing a knitted wool hat and scarf, cyberpunk style
How to Get Started
Via ChatGPT
Accessing DALL·E 3 through ChatGPT is straightforward.
Go to https://chat.openai.com/ – you will need to be on the paid plan and use ChatGPT-4
Start a ChatGPT session and use the ‘message ChatGPT’ field to describe the image you want to have generated.
Via MS Copilot/Designer
Login to your Microsoft account.
Go to: https://copilot.microsoft.com/
Use the ‘Ask me anything…’ field to describe the image you want to have generated.
In designer you have several options to customize your design. If you are familiar with Canva, the Designer can feel easy to navigate.
Bottom Line
DALL·E 3 serves as an easy entry point to AI image generation, especially if you are already using ChatGPT or Microsoft ecosystem.
Freepik
Yes, the stock image website. Freepik may not be an obvious option for AI image generation, but they have recently added impressive selection of AI image tools. They have one of the, if not the most beginner-friendly user interface making it super easy to get started.
Super easy and intuitive to use with customization options. Great for photorealistic images for your website, blog, social media, etc.
If you want to create highly artistic, creative images with custom controls, Freepik may not be the best option for you.
SPECS
Input method: text prompt
Default image size: 1024×1024
Aspect ratio: square (default) + 6 other options available
Output format: JPEG
Number of images generated per run: infinite
Access: website, mobile app
Generation time: immediate
Prompting knowledge: helpful
Other features: Sketch-to-image, Enhance, Reimagine, HD download, advanced features
PROS
Ease of use: Easy
Infinite scroll: generates unlimited images from a single prompt
Super fast: immediate image generation
Excels at generating photorealistic images
Advanced features: several
HD download option (4096×4096)
Full license to use any assets commercially (paid plan)
CONS
Falls short in generating highly creative images
Prompt adherence: Average
PRICING
Free Plan
20 daily AI text-to-image generations. Limited use of AI Art generator with Pikaso.
Paid Plan
Premium: $144/year ($12/month) or $24 for one month.
Unlimited AI image generations. All other Freepik’s offerings are included (163M premium vectors, photos, AI images, icons, videos, PSD… and Flaticon). The plan also includes a full license to use any asset commercially.
Please read their full documentation on usage rights. AI images copyright and usage rules
Examples
Prompt:
Futuristic city skyline at night, with a large illuminated building and a river sparkling below
A photo of a delicious cheeseburger
Prompt:
a photograph of a blond woman wearing a knitted wool hat and scarf, cyberpunk style
How to Get Started
Go to: Freepik.com > text-to-image generation
In the field at the top of the page labeled: describe what you want to see, describe what you want to create.
Click create.
Make any adjustment you want from style, color, aspect ratio and more.
Save and/or download the images you want to keep.
If you want to change an image – try Reimagine.
Bottom Line
Freepik AI image generation tool is a fantastic one to start with. The ease of use, infinite scroll, fast – real time image generation and premade style and other features make it super easy for beginners.
Freepik does a great job with photorealistic images, however if you want a tool that is more artistic – it’s next.
Midjourney
Although Midjourney is not the easiest AI image generator to figure out, I need to include it here for it’s outstanding quality and endless variety of images of all types it can generate.
Midjourney is an AI art exploration, for some even art therapy. You will be frequently delighted and at awe of the images Midjourney can generate.
Offers you fantastic creative exploration and control with outstanding quality. Midjourney is what many think of when talking about AI art.
Currently only accessible via Discord, which can be challenging to figure out how to use. (website will be soon available)
SPECS
Input method: text prompt + advanced options
Default image size: 1024×1024
Aspect ratio: square (default) + several others to choose from
Output format: PNG
Number of images generated per run: 4
Access: Discord (website, mobile – coming soon)
Generation time: fast, a few seconds (on fast generation), 3 modes: relaxed, fast, turbo
Prompting knowledge: recommended
Other features: upscale, style, image and character references, variations, custom adjustments, inpainting, more.
PROS
Fantastic image quality – all styles
Many image customization and modification options
Great creative control
Several advanced features
Good prompt adherence
Can generate text into an image
Easy to view and organize your images (archive)
CONS
Only accessible via Discord – can be hard to figure out
Requires at least some understanding of the variety of levers and parameters for advanced image generation
PRICING
No free plan.
Paid Plans
Starting at $10/month (yearly subscription $96/yr = $8/mo).
If you run out of your hours, you can pay for additional hours – $4/hr.
Starting at Pro level plan ($60/month) you can generate in ‘stealth’ mode, where your images are only visible to you.
You can use ‘relaxed’, ‘fast’ or ‘turbo’ modes. Relaxed mode (slower) in any paid level does not eat up your hours. Meaning you can generate as many images as you want in relaxed mode.
Examples
Prompt:
Futuristic city skyline at night, with a large illuminated building and a river sparkling below
A photo of a delicious cheeseburger
Prompt:
a photograph of a blond woman wearing a knitted wool hat and scarf, cyberpunk style
How to Get Started
Follow the directions here (Midjourney documentation).
Prior Discord knowledge is helpful.
If you find the navigation to be difficult, you are not the only one. Not to worry though, the new website with image generation is going to be available very soon. As soon as it is – I will update this article and add a video tutorial.
Bottom Line
Midjourney can be challenging to get started with, but it offers such fantastic and versatile creative exploration and creation of images with outstanding quality it’s hard to beat. It’s one of the top AI image generators available.
Good to Know
Each AI image generator has technical limitations. For example, at times you can get artifacts, extra fingers, or similar, in your images. Or some models don’t understand the prompts as well as others, creating images that do not accurately match what you asked it to create. It is the current state of the image generators.
Some image generators have visible watermarks on the images. This is more typical practice for free plans.
Understanding prompting is hugely useful in improving your images.
You do NOT have to pay for prompts!! There are plenty of prompts out there. I also share prompts you can take and use. Also keep in mind that not all prompts are created equal – even with paid prompts. The more you learn about prompting, the better prompts you can create.
If image usage rights are important to you, check the TOS of each platform and consult an attorney to determine what is and what is not allowed.
NOTE
All tools or devices may not be available for your country/region at this time. Check for availability.