The explosion of AI that generates images from text has made creating images faster than ever before. AI image generator brings your artistic vision to life by converting simple text prompts into corresponding images. Whether you’re an AI enthusiast looking to try out a new program or an amateur creator looking for help, AI Generator can help you brainstorm, generate new ideas, and mock up potential projects. and helps you quickly create personalized art. Choosing the right AI image generator for you is important, but choosing from all the different options available can be overwhelming. Not all AI image generators are created equal, and not all services offer the tools or protection you need. That’s why CNET has spent a lot of time testing and reviewing different services to find out which one is best for you.
The AI image generators on this list have undergone rigorous testing. CNET reviewers spent months with these programs and generated hundreds of images, creating everything from cartoon rainbow safaris to dramatic sci-fi scenes to photorealistic stock images. During these tests, all services on this list spit out unstable or unusable images. Even the best isn’t perfect. The sign of a truly great AI image generator is how well equipped it is to handle its quirks and fix its flaws. Editing tools and customization options are a big part of it, so test them extensively. Privacy policies, such as how the generator trains its content, are also important considerations when choosing the best service. This was the deciding factor.
Read more: How to create AI images: A complete guide with expert advice
What is the best AI image generator?
OpenAI’s Dall-E 3 is our pick for the best AI image generator. It can handle long and complex queries, offers many editing and customization options, and has a unique conversation flow that makes it easy to realize your artistic vision. If you don’t want to sign up for a $20/month ChatGPT Plus subscription, Leonardo AI’s free plan offers great AI images. Beginners will find Canva’s user-friendly design easy to use and use. Professional creators looking to experiment with AI art should try Adobe Firefly and its family of AI tools.
Best AI Image Generator of 2024
like
Understand long and complex queries Attractive and dynamic images Conversational and easy to modify
I don’t like it
Photorealistic images may look fake Image generation is slow
OpenAI’s Dall-E 3 was named CNET’s 2024 Editors’ Choice Best AI Image Generator. Dall-E 3 is perfect for amateurs and professionals alike, as it offers many editing tools, including different styles and customizations. Dall-E is unique among its competitors for its conversation flow, ease of sending follow-up requests, and ability to handle complex queries. While OpenAI may use your content to improve its models, the company said it will not share your data with advertisers and will provide a way to request to opt out of training your models. are. With a $20/month ChatGPT Plus subscription, Dall-E 3 creates vibrant, engaging images with less AI quirks.
Read our review of Dall-E 3.
like
Clear, detailed images Helpful, rapid engineering tools
I don’t like it
Free plan has no post-generation editing tools Lackluster privacy policy
Leonardo is perfect for AI creatives on a budget. It has a comprehensive free plan and provides ample generating credits at high speed. There are no free post-generation editing tools (you have to pay), but there are prompt improvement tools and other customization elements to help you build the best prompts. Leonardo makes it easy to explore what AI can do and create crisp, detailed AI images. Its free plan easily beats out competitors, even the user-friendly Canva.
Read our review of Leonardo AI.
like
Great artistic style Great tweaking options Competitive generation speed
I don’t like it
Struggles with photorealistic images Difficulty when elements touch or interact Does not accept complex queries
Adobe Firefly’s family of generated AI image tools is built directly into Adobe Creative Cloud, including Photoshop, making it a great option for professional creatives looking to experiment. Firefly offers many stylistic and artistic options, and its adjustment tools are similar to the editing software creators are used to. Firefly is trained on Adobe’s own Stock catalog of high-quality licensed public domain content. Firefly does not train content. If you’re already paying for an Adobe Creative Cloud subscription, Firefly is an easy way to mock up ideas and spark inspiration.
Read our review of Adobe Firefly.
like
Great for creating fun, quirky, cartoony images Easy to use on desktop and mobile apps We don’t share your content to train our AI
I don’t like it
The free plan has severe limits on the number of images you can create Difficult to generate non-square images
Canva is a very user-friendly graphic design service, and its AI image generation tool Magic Media is a great choice for amateur AI creators. Magic Media is a minimalist service, so it’s not a good choice if you need extensive editing tools, but it’s great for people with limited budget and time. Unlike many competitors, Canva doesn’t train its AI on your content and the images it generates are always private, so Canva’s privacy policy is particularly secure. Canva also makes it easy to integrate AI images into other projects on desktop and mobile apps. A no-frills, easy-to-use AI image generator perfect for beginners and Canva enthusiasts.
Read our Canva review.
How we tested our AI image generator
CNET takes a hands-on approach to reviewing AI image generators. Our goal is to determine how well it compares to its competitors and for which purposes it is best suited. To do this, provide the AI with prompts based on real-world use cases, such as rendering in a certain style, combining elements into a single image, or handling longer descriptions. We score image generators on a 10-point scale, taking into account factors such as how well the images match the prompt, the creativity of the results, and the speed of response. Learn more about how to test AI.
Factors to consider
Accuracy is an important factor to consider when evaluating AI image generators. Accuracy in this case refers to how well the service matches the prompt and how clearly elements and details are displayed. For example, a service that closely matches the sci-fi/fantasy spaceship scene described in the prompt will be considered accurate. The same goes for services that generate images of humans with clear facial expressions.
On the flip side, it’s also important to consider hallucinations and how AI can fabricate things. Hallucinations in AI images tend to be strange quirks and flaws that you didn’t ask for, like figures with three arms or elements disappearing. AI image generators inherently require creativity, but hallucinations are a clear failure, not a quirky or unrealistic style element. All AI image generators will encounter these issues, but good AI services will encounter them less frequently and have editing tools to fix them.
The number of disambiguation prompts required indicates how much work you need to expend to get the image you want. If you can’t follow up on requests for edits or additions, this could be a red flag or annoyance that you need to be aware of. A generator that closely follows prompts and provides editing tools will help you bring your vision to life.
It is also important to consider response speed. Most image generators are very fast, taking around 2 minutes. Services that can generate images 10 to 30 seconds faster than that provide a better user experience.
Other AI image generators we tested
Midjourney is a solid choice for an AI image generator, but it’s currently only available on Discord, is paid-only, and inconsistently matches prompts, which kept it from being a top pick. Midjourney offers great upscaling and editing tools for individual images, but you’ll need to use them frequently. It’s also worth noting that all images are public and accessible in your online gallery unless you create them in stealth mode. Stealth mode is only available on the more expensive Pro and Mega plans.
See here: How to use ChatGPT’s new voice conversations
04:27
Google’s ImageFX also wasn’t a top choice due to repeated inaccuracies and “overbearing babysitting” that results in rejecting innocuous prompts. All AI generators struggle with photo-realistic faces and hands, but ImageFX also struggles with inanimate objects and logos. This isn’t a big problem, except that the editing and tweaking tools aren’t quite up to the point where you can use the image.
OpenAI’s Dall-E 3 was chosen by CNET editors as the best AI image generator for its wide range of editing and customization tools, strict adherence to prompts, and level of image detail. Dall-E is great at using conversational flow and handling more complex prompts appropriately.
See more
There are several free and freemium AI image generators to choose from, but we recommend starting with Leonardo AI or Canva. Both of these services offer free AI image generation with fast response speeds and good instant compliance.
See more
To create the best AI images, you need to build good image prompts. When creating what you want to create with the generator, make sure it includes the style, aesthetics, dimensions, and specific elements you want to include. Certain prompts can be difficult to follow depending on the generator, but it’s a good idea to start with specific prompts and adjust or scale back as needed.
See more
There are many reasons why an AI image generator may not provide the desired results. Try using customization settings and editing tools to adjust your prompts to eliminate quirks and conflicting elements. Unfortunately, sometimes your only option is to scrap your idea and start over.
See more
yes. If you are using an image created entirely by an AI image generator, you must disclose that wherever you share it.
See more
