The best Adobe Firefly alternatives for making and editing AI images, including Midjourney, Leonardo, ChatGPT Images, Runway and Google's Nano Banana.
Some links in this post are affiliate links. If you buy through one, I may earn a commission at no extra cost to you.
The short answer
Midjourney is still the best Adobe Firefly alternative for image quality, and its web Editor now does Firefly-style generative fill on your own photos too. Leonardo is the best value, with a free plan and a Canvas editor. For quick edits from a text prompt, ChatGPT Images and Google Gemini's Nano Banana are the easiest picks.
I've spent a LOT of time playing with the various AI image generation tools over the years.
When Adobe Firefly was announced in early 2023, the generative fill function caught the attention of creators across the globe.
In this guide, I've put together the best Adobe Firefly alternatives that feature generative fill-style editing.
Midjourney is an industry favorite for image quality, and its web Editor does generative fill (Vary Region) on Midjourney images and your own uploads.
Leonardo offers its own image models plus partner models, a prompt-based Inline Editor, and a Canvas editor for inpainting and outpainting.
Built right into ChatGPT on every plan, ChatGPT Images can make images and edit any part of them, including your own photos, from a simple prompt.
Runway is the pick for video. Its Aleph model adds, removes and changes things in footage from a text prompt, and it makes images too.
Google's Nano Banana image model lives inside Gemini. Upload a photo, describe the change, and it edits it for you. The base model is free.
These were the main considerations I took when determining the best Adobe Firefly alternatives!
Even though Adobe Firefly is one of the top AI image generation tools on the market, there's a number of reasons creators are looking to use different tools to create images.
This is perhaps the most common reason creators are looking to move away from Adobe Firefly. When I tested them, I found options like Midjourney, Leonardo, and Dall-E 3 much better at creating images. With Midjourney's Vary (Region) and Editor tools, there isn't much reason to choose Firefly when it comes to quality. To be fair, Firefly now includes partner models like Google's Nano Banana and OpenAI's GPT Image too, which narrows the gap.
To use Adobe Firefly you must first sign up and create an account with Adobe. Although it's not a huge hassle, it can put off new users who might just be looking to try out the tool and see if they like it.
Unlike most of its competitors, Adobe trains its own Firefly models only on licensed and public domain content. While this is great for users who might be worried about copyright implications, it has resulted in the overall variety of the output lacking quite a bit compared to its competitors. There is also quite a heavy content filter on the images produced with Firefly, so prompts requesting images of celebrities and other potentially controversial topics do not work.
Adobe Firefly is a generative AI tool released in 2023 that can create custom images from a natural language input, alongside a range of other automated image generation features.
These days it's a full creative app for images, video and audio. Adobe's own Firefly Image 5 model sits alongside partner models from Google, OpenAI, Black Forest Labs (FLUX), Runway and others, and it works on the web and as a mobile app for iOS and Android (Adobe).
Upon launch, Firefly attracted attention for its generative fill functionality which allowed users to insert, modify, and remove items from images with simple text prompts.
Firefly Generative Fill (GIF: Adobe)
Firefly has all the features we've come to love from AI image generation tools: text-to-image, Generative Fill, Generative Expand for extending an image, and Prompt to Edit, which changes an image from a written instruction.
Early on, Firefly also had text effects, which was handy because most AI generated images back then struggled with lettering and correct spelling.
Firefly Image Extension (GIF: Adobe)
Adobe has built Firefly with ethical considerations in mind, and it pays Adobe Stock contributors whose content was used to train its models (Adobe).
Adobe's own Firefly models are trained on licensed content like Adobe Stock plus public domain images rather than content scraped from the web, and Adobe designed them to be commercially safe. Enterprise customers even get IP indemnification. This is a massive selling point for companies looking to avoid legal issues associated with AI image generation. Just note that the partner models inside Firefly aren't covered by the same promise, and Adobe leaves it up to you to decide whether they suit your project.
The addition of Firefly into Photoshop is a game changer for users allowing massive time savings and making the barrier to entry for using Photoshop much, much lower. Photoshop's Generative Fill even lets you pick partner models like Google's Nano Banana Pro and FLUX.2 pro on desktop (Adobe).
Adobe's plans and prices have changed since I took this screenshot.
The best Adobe Firefly alternatives are Midjourney, Leonardo, ChatGPT Images, Runway and Google Gemini. While Firefly is an excellent generative AI tool for images and video, that doesn't mean it's the best solution for everyone.
Here are the top 5 Adobe Firefly Alternatives:
Midjourney is one of the most popular image generation tools available right now – and for good reason.
The image quality from Midjourney is simply incredible. Users who are new to generative AI can expect to be amazed at the art generation capabilities of Midjourney, I know I was.
Midjourney can generate basically anything you can think of whether it be ultra-real stock photos, product logos, cartoon characters, or professional food shots. If you can imagine it, Midjourney can do it!
Created using Midjourney v5.1
Another massive benefit of Midjourney is that due to its popularity, there are heaps of online resources to help you craft the best prompts, like my Midjourney cheat sheet.
Midjourney also has its very own generative fill type function known as Vary (Region), and it's extremely good.
Check out the image below. I generated it with Midjourney, then added a neck tattoo with the Vary (Region) feature. The quality certainly gives Firefly a run for its money.
I created this image in Midjourney, and then used the Vary(Region) function to add a neck tattoo to the lady.
Midjourney has all the standard image generation modes like text-to-image, text + image-to-image, image overlaying, and even reverse image prompting which allows you to upload an image and generate a text prompt.
Midjourney works on its website, midjourney.com, as well as in Discord. The web Editor is where the Firefly-style tools live: you can erase and regenerate parts of an image (Vary Region), pan, zoom out, and even upload your own photos to edit. The current V8.2 model, the default since July 2026, also brought an Edit model that changes images from written instructions (Midjourney).
Given the extremely high-quality images provided by Midjourney, and its own type of generative fill functionality called Vary (Region), it's still my pick for the best Adobe Firefly alternative.
Created using Midjourney v5.1
Leonardo provides AI image generation with an impressive range of features and capabilities. It's now owned by Canva but still runs as its own product (Canva).
Alongside its own models like Phoenix and Lucid Origin, Leonardo gives you partner models such as Nano Banana, GPT Image and FLUX inside the same app, so you can generate images in a huge range of styles.
Leonardo supports different types of prompt types for image generation, including text, image prompts, and image-to-image.
When I first tested it, Leonardo's Alchemy and Prompt Magic v2 features made impressive 2D images, like this one. 👇
Once your image generation is complete, Leonardo allows you to refine your creation further with its post-processing features.
Upscaling, unzooming, background removal, and transfer to the Canvas are all options available to you for enhanced control and post processing of your generated images.
Leonardo's Canvas Editor fully qualifies it as one of the Firefly alternatives. Extend, mask, replace, and erase elements indefinitely and effortlessly, and regenerate specific areas to achieve your desired output.
These days there's also an Inline Editor: open any image, type what you want changed, and Leonardo's Omni models make the edit, with up to six reference images depending on the model.
Alongside its ability to generate images, I was quite impressed with Leonardo's inpaint/outpaint functionality. See below for an example where I removed the glasses from the lady.
An example I generated to remove the women's glasses using Leonardo's take on generative fill.
Leonardo also features an LLM for automatic prompt generation describing your digital art, which I thought was a neat feature.
You can also train your own custom models (Leonardo calls them Elements) on your own images, and Leonardo has branched out into video and 3D generation too.
I've covered every tool in depth in my Leonardo AI review.
Leonardo is a comprehensive AI art generator that offers a seamless design process and control over your creations, making it a competitive choice in my list of Firefly alternatives in second place.
Leonardo has renamed and repriced its plans since I took this screenshot.
Leonardo offers 4 pricing tiers:
ChatGPT Images is OpenAI's image generator, built right into ChatGPT. It replaced DALL·E 3, which is what I originally tested for this list.
OpenAI has since retired DALL·E 2 and DALL·E 3 altogether, and their API access ended in May 2026 (OpenAI).
The current version, ChatGPT Images 2.5, can create images from a prompt and edit them. Select part of an image with the selection tool and describe the change, or just describe the edit in chat. It works on your own uploaded photos too, which makes it a proper generative fill alternative (OpenAI).
Back when I tested DALL·E, outpainting was only available in the Dall-E web app (not ChatGPT).
To get a feel for Dall-E's outpainting, I took an image of a puppy and used Dall-E to give it a beanie. The results weren't bad!
An example of Dall-E's Inpainting functionality with a Midjourney puppy (left) and Dall-E Outpainted 'beanie' (right).
The image quality from Dall-E 3 was already significantly better than what we'd seen in Dall-E 2. Check out the two images below where I used the same prompt in Dall-E 2 and Dall-E 3 to compare the difference.
An example of Dall-E 2 Vs. Dall-E 3 puppies!
Another thing I noticed right away with Dall-E 3 was strict restrictions on specific words. For example, I wasn't able to use the word "Disney" in my prompt, assumably due to fears of copyright issues.
Old screenshot: DALL·E 2's credit packs are long gone.
Runway is an AI art generator that boasts the widest range of capabilities of any of the tools I tried.
When I first tested it, Runway's flagship was Gen-2. It's now on Gen-4.5 for video generation, plus Aleph for editing existing footage.
Runway has earned the most notoriety for its generative AI video content capabilities and features a video-to-video tool which allows you to take existing footage and change the style – the clip below shows what you can do with Runway's video-to-video feature.
Video-to-video using Runway Gen-1 (GIF: Runway)
Runway's AI video features can do a lot more than video-to-video now. Its Aleph model can add, remove, and transform objects in a video, generate new camera angles, and change the style and lighting of a shot, all from a text prompt (Runway). Aleph 2.0 goes further: edit one frame and Runway applies the change to the rest of the clip.
On top of the video features, Runway makes images with its Gen-4 Image model and gives you partner image models like Nano Banana Pro, as well as music, voice, and sound effects generation.
When I tested it, the biggest drawback of Runway was image quality, which lagged behind tools that focused on one area of generative AI only. Runway now hosts partner image models too, so that's less of an issue.
Given its huge range of tools and the strongest video editing on this list, Runway holds fourth place as an Adobe Firefly alternative.
Runway's plans have changed since I took this screenshot.
Google Gemini has become one of the easiest ways to edit images with AI, thanks to Google's image model, Nano Banana.
Upload a photo, tell Gemini what to change, and it makes the edit. You can swap a sunny day for a moody night, change the camera angle, restyle the whole image from a reference photo, or combine two photos into one (Google).
It's also good with text, so logos, posters, and invites come out with clear, readable words in many languages.
Nano Banana 2 is free in the Gemini app, and Google AI plans unlock Nano Banana 2.1, Google's best image model, with higher limits.
Adobe even offers Google's Nano Banana models inside Firefly and Photoshop's Generative Fill, which tells you how far it's come.
If you want Firefly-style edits for free, with nothing more than a text prompt, Gemini is a great place to start.
Anyone who wants AI images or Firefly-style edits without an Adobe subscription, or who wants better results than Firefly gives them, should try these alternatives.
Generative AI image and video tools can be used by creatives looking to come up with ideas and spark creativity, businesses looking to save time and money on image or logo generation, Freelancers looking to sell their creations, researchers looking to quickly and easily visualize concepts for their audience, teachers can use them to create custom learning materials, entertainers can create realistic environments and characters, the possibilities are endless.
The best part about these tools is how easily accessible and user-friendly most of them are. With just a few clicks of your mouse, you can be using cutting-edge technology and seriously level up your productivity and creativity.
If you're mainly after image generation, my list of Midjourney alternatives covers even more options.
Adobe Firefly brought a lot of attention to the generative AI space with its high-quality generative fill feature when it launched in 2023.
As you can see in the examples I provided in this article, the tools each have unique strengths and weaknesses that can compete with Firefly.
Midjourney is my choice for the best Firefly alternative due to the extremely high quality outputs and versatility of its Vary (Region) functionality. Leonardo is a close second.
Midjourney makes some of the highest quality AI images around, with the widest range of styles. Its web Editor and Vary Region feature make Midjourney my pick for the best Firefly alternative.
Questions people ask
Yes, partly. Firefly has a free plan with a daily allowance of generations, and paid plans (Standard, Pro, Pro Plus and Premium) add 2,000 to 50,000 generative credits a month. See Adobe's Firefly plans for prices in your currency.
Leonardo or Google Gemini. Leonardo's free plan gives you 150 tokens a day plus its editing tools, and Gemini's free Nano Banana 2 model can generate and edit images from a simple prompt.
Yes. Midjourney's Editor on midjourney.com lets you erase part of an image and regenerate it with a prompt (Vary Region), extend the canvas, and edit your own uploaded photos. Its V8 Edit model can also change images from written instructions.
Yes. In ChatGPT Images you can highlight part of an image with the selection tool and describe what to change, or just describe the edit in chat. It works on images ChatGPT made and on photos you upload.
$10 a month for the Basic plan. Standard is $30, Pro is $60 and Mega is $120 a month, and paying yearly takes 20% off. There's no free plan.
Keep reading
Updated Oct 14, 2026 · 11 min read
Updated Oct 9, 2026 · 4 min read