The Top 3 AI Image Generators to Try in 2024
AI image generators use text prompts to make pictures very quickly. I looked at the best AI image generators and found that there is not only one option that you can use. Here are the top 3 AI image generators to try in 2024.
1. MidJourney:
The results from Midjourney are always my favourite of all the picture generators on this list. Along with better colours and textures, the pictures it makes look more cohesive and are more interesting to look at overall. People and real-world things look more natural and like real life than they do with other AI image generators, at least when they aren’t told to. The newest versions can even get hands kind of right. This was the first AI-made picture that won an art contest, which says a lot.
For now, Midjourney is let down by a big bug: you can only get to the beta through Discord. If you join Midjourney’s Discord server or ask the Midjourney bot on a server you run, you can type /imagine [whatever you want to see] to get to a prompt. The bot will then come up with four different versions of your question, which you can download, print out, change, and do other things with.
This will soon change, which is good news. A real web app is being beta tested by Midjourney. The company says it will continue to support Discord for a long time. There may still be some bugs, but it’s great that the best AI picture generator will soon be for more people.
Regarding those quirks, every picture you make is automatically shared with everyone in Midjourney’s Discord and can be seen on your website page. Not only does it make things feel more like a group, but anyone can see what you’re making. This might not be a big deal for artists, but it could be a deal-breaker if you want to use Midjourney for business.
Do not worry if all of this sounds hard to understand. The Midjourney help files are very helpful. They explain how to get started and also all of the more advanced features, such as the different model versions, upscaling your pictures, blending multiple images, and controlling things with different parameters. The results you can get are truly amazing once you know what your choices are.
Pros
- The group is a great place to find ideas.
- Always makes the AI-generated pictures that look the best.
Cons
- Trials for free are currently on hold
- It’s weird that it can only be used through Discord.
Pricing:
Starting at $10 a month for the Basic Plan, which lets you make about 200 pictures a month and gives you the right to use them commercially.
2. DALL-E 3:
There’s a good reason why DALL·E 3 is one of the most well-known AI picture generators. Its predecessor, DALL·E 2, was the first AI-powered image maker that could make highly interesting pictures and was used by enough people for them to go viral.
You can still use DALL·E 2 through its web app and API, but DALL·E 3 is much better and can be used through ChatGPT or Microsoft Bing’s AI Copilot. For every challenge, it leads to more interesting, realistic, and consistent writing. With AI image generators, OpenAI seemed to be falling behind its rivals before DALL·E 3. But now it’s right back in the race.
The best thing about DALL·E 3 is how ridiculously easy it is to use. Just tell ChatGPT or Bing what you want to see, and they’ll give you two to four different versions to choose from right away. It uses GPT-4’s knowledge of language to give you more prompts. This way, each answer will be different, and you can always ask for more.
There is no longer a free way to try DALL·E 3 from OpenAI, but Microsoft does. I had a harder time with Copilot because it wouldn’t change based on follow-up hints, but the price is hard to beat.
ChatGPT is the best way to use DALL·E 3, though. If you pay for ChatGPT Plus, you can use it as much as you want, but GPT-4 only lets you send 40 texts every three hours.
You can edit photos in DALL·E 3 in two ways: you can ask ChatGPT to make changes, and it will restart the prompts with your changes; or you can use a select tool to make changes only happen in certain parts of the picture. It’s not possible to make a DALL·E picture bigger yet, at least not without switching to a different app. It can feel like magic when ChatGPT does exactly what you ask it to do. It’s sometimes like working with an intern who is too eager to please and too happy to do their own thing. You can try one of these other apps if you want more power.
Pros
- You get a lot of AI for your money with ChatGPT Plus because it comes with it.
- Really simple to use
Cons
- It’s expensive at $20 a month if you don’t want GPT.
- Your ChatGPT settings aren’t always reliable.
Pricing:
DALL·E 3 is free through Microsoft Copilot and part of ChatGPT Plus, which costs $20 a month. DALL·E 2 costs $15 for 115 picture credits. Prices for APIs are more complicated, but images start at $0.016 each.
3. Adobe Firefly:
It shouldn’t be a wonder that Adobe has one of the best text-to-image generators, since it has been adding AI tools to its apps for more than 15 years. At least, it works well with other Adobe apps. Firefly is its AI model. You can try it out for free on the web or through Adobe Express, but the newest version of Photoshop is where it really shines.
Firefly has some hidden skills. It can not only make new images from a detailed text description, but it can also change the colour of vector art, add AI-generated elements to your images, and make text effects from a written prompt. The web app lets you try all of these out, but the last one is what makes Firefly stand out.
If you only use Firefly to turn text into images, the effects aren’t always great. A lot of the time, it can match or beat DALL·E, or stable diffusion. But sometimes, I’m not sure what it was trying to do. But the fact that it works with Photoshop, which is the standard picture editor, is really cool.
The name of this function is generative fill. The idea is that you can use Photoshop’s normal tools to pick out a part of your picture and then add something else by clicking a button and typing a word. Most importantly, Generative Fill knows what your picture is about. You can see that Photoshop has matched the blurred depth of field for the trees I added with Generative Fill in the picture above. It looks like a whole.
Adobe’s Firefly is the first AI picture generator that really shows what’s to come, even though DALL·E and Stable Diffusion got people talking about AIs that can make images. This is not a trick; it’s a tool that millions of workers who use Adobe apps every day can use.
Pros
- Style and Structure Guide
- Free to use
- Safe for businesses
Cons
- Needs more specific prompts
- This generator has a longer lag than others.
Pricing:
You can get 25 credits for free, or 100 credits for $4.99 a month. Photoshop costs $19.99 a month as part of the Creative Cloud Photography Plan, which gives you 500 creative credits
