I have tried many AI tools, and I have chosen Midjourney because of the quality of its images. I mainly use it to make images for blog posts and web pages. About 90% of the time, the output is close to what I have in mind. It also has a feature where you give it a prompt and get four variations of an image. You choose one, and it creates more variations in that direction. This helps you broaden your idea and add more value to the image you create. Review collected by and hosted on G2.com.

I don’t have a major dislike, but a few things could be improved. First, the text inside images is mostly accurate in English, but when I switch to a regional language like Hindi or Gujarati, it starts getting it wrong (30% of the time). Second, when I am working through one flow and have already described how I want the image to look, adding something in between can sometimes pull the result in a completely different direction, as if it lost the earlier details. This is not frequent, but it does happen sometimes. Review collected by and hosted on G2.com.

What I like most about Midjourney is how it can generate stunning, high-quality images from even simple text prompts. Its creativity, strong attention to detail, realistic textures, and broad range of artistic styles make it a powerful tool for bringing ideas to life. I find it especially useful for concept art and design inspiration, and for quickly exploring different visual possibilities. Review collected by and hosted on G2.com.

What I dislike about Midjourney is that it can sometimes struggle to follow complex prompts accurately, especially when I’m trying to generate specific details, readable text, or a very precise composition. It often takes some trial and error—and a few prompt tweaks—to get the results I’m aiming for. On top of that, the learning curve can feel steep for beginners, and the subscription cost may be a drawback if you’re only a casual user. Review collected by and hosted on G2.com.

what i like most about midjourney is being able to experiment with different ideas without having to spend a ton of time actually creating the images from scratch. ive used it mostly for image creation and i like how you can start with a pretty bsaic prompt, see the differnent options and then adjust further. the image variations are probably my fav part because sometimes ut comes up with something that i never envisioned but in a really good way. its also very useful to see different concepts without needing or learning things like adobe photoshop. Review collected by and hosted on G2.com.

one thing that i didnt like initially is that i can take a few attempts to get an image that i actually imagined in my head. sometimes i feel like my prompt is pretty clear, but the end result is a little off. ive also found that figuring out how to write effective prompts can take some getting used to. it was a bit tedious and i wished (atleast in the beginning) there was a more straightforward way to make smaller changes without having to redo the entire image Review collected by and hosted on G2.com.

What I like most is how quickly I can go from a rough idea to something I’d actually use. I mostly create concept art for a side project, and that first batch of four images usually gets me about 70% of the way there. Having consistent style across prompts saves me a lot of rework and makes it easier to keep things cohesive. The pricing feels fair for how much I generate, especially compared with paying for stock images. I also like the Discord and web workflow, since it lets me iterate without much setup. The community gallery ended up being more helpful than I expected, too, especially when I was learning how to phrase prompts. Review collected by and hosted on G2.com.

What I like least is how much the workflow still depends on Discord. Scrolling through a busy channel to find an image I generated last week is tedious, and while the web app has improved things, it still doesn’t fully replace that experience. My biggest frustration is the pricing: once I used up my Fast GPU hours [mid-month / in about X days], the Relax queue made every batch feel slow. Text inside images also still comes out garbled more often than I’d like, which is especially frustrating when I need something like a logo or a label. I'd also like a public API so I can connect it to my own tools. Finally, the first few weeks took longer than they should have because the prompt syntax isn’t explained very well. Review collected by and hosted on G2.com.

What I like most about Midjourney is that I can have an idea in my head and actually see it come to life without needing to be good at drawing. I enjoy playing around with different prompts and seeing how the same idea can turn out in completely different ways. The image quality is usually impressive, especially when it comes to lighting and small details. Of course, it doesn't always get things right on the first try, and sometimes I have to experiment quite a bit to get what I'm looking for. But that's also what makes it fun to use. It gives me the freedom to try things out and be creative without needing a lot of technical skills. Review collected by and hosted on G2.com.

The biggest thing that frustrates me about Midjourney is that getting a very specific result can take more attempts than I expect. I might have a clear idea of what I want, but even after adjusting the prompt, certain details like hand positions, facial expressions, or the placement of objects don't always come out right. Sometimes the overall image looks great, but one small detail makes it unusable for what I had in mind. I also find that changing one part of a prompt can affect other parts of the image that I wanted to keep. It's a great tool for experimenting, but when I need more precise control over the final result, the trial-and-error process can get tiring. Review collected by and hosted on G2.com.

I really like how well it supports exploration. Being able to generate a grid of variations and then remix, vary, or upscale the ones that stand out matches how visual ideas actually develop - you often don’t know what you want until you’ve seen a few options. The style references, along with the ability to carry a consistent look across images, make that iterative process feel more practical and genuinely useful for real projects. Review collected by and hosted on G2.com.

The flip side of its strong aesthetic is my biggest gripe: Midjourney can be so opinionated that it sometimes overrides what you actually asked for. The images often drift toward a polished, glossy, slightly dramatic look, so trying to get something intentionally plain, awkward, or documentary-style can take a lot of wrangling. It also tends to lag some competitors when it comes to precise prompt adherence. For example, hitting exact object counts, nailing specific spatial arrangements, or producing legible text within the image. Review collected by and hosted on G2.com.

Unlocks Creative Confidence: Enables non-technical team members to convert complex hardware concepts into professional visuals without 3D skills.

Streamlines Collaboration: Functions as a fast, shared visual language for team brainstorming, client pitches, CMF moodboards, and storyboards.

Intelligent Editing & Continuity: Uses Character/Style References and Selective Inpainting to refine specific details and maintain brand consistency across campaigns.

Massive Time Savings: Cuts days of stock searching and rough drafting down to minutes, accelerating campaign launches.

Total Creative Freedom: Delivers high-end photorealism and custom artistic styles tailored to any brand or format.

Integration & Onboarding: Fits smoothly into Adobe Creative Cloud for final layouts, backed by an intuitive Web UI for fast team onboarding. Review collected by and hosted on G2.com.

Text Rendering Issues: Struggles to generate accurate text, exact spellings, or clean vector logos within images.

Lack of Precise Editing Control: Doesn't allow manual pixel or vector editing, requiring trial-and-error re-prompts for small fixes.

Anatomical Glitches: Occasionally creates subtle errors like extra fingers, distorted limbs, or unnatural floating objects.

Prompting Learning Curve: Getting a specific output requires mastering technical parameters and trial-and-error prompt tweaking. Review collected by and hosted on G2.com.

The image quality and aesthetic range are what keep me coming back. As a designer working on brand and marketing assets, I use Midjourney mostly for early concepting: mood boards, art direction explorations, and background textures I can refine in Photoshop or bring into motion work. The style reference and image prompt features make it easy to stay close to an existing visual direction instead of starting from scratch every time. Variations and upscaling let me iterate quickly on one strong idea. The web editor has also made it much easier to tweak compositions without bouncing between tools. Overall, it cuts the time from rough idea to something I can show a stakeholder from hours down to minutes. Review collected by and hosted on G2.com.

The biggest frustration is precise control. Midjourney is great for mood and atmosphere, but it’s hard to get exact details right, like specific text, logos, accurate product shapes, or a precise layout. That means most outputs need cleanup in Photoshop before they’re usable for brand work. There are also no real integrations with the tools I use every day, like Figma, Adobe, or a shared asset library, so a lot of time goes to downloading and re-uploading files. Generation speed can slow down during peak hours, and fast GPU hours run out quickly on heavier projects. Licensing and usage terms could also be clearer for commercial and team use. Native integrations and better control over specific elements would make it much easier to use in a production workflow. Review collected by and hosted on G2.com.

I really like how easy it is to turn an idea into something visual with Midjourney. For example, if I want to picture a cozy living room with warm colors and modern furniture, I can just describe it and get some really nice results in seconds. I also like trying different styles and playing around with lighting and details without having to start from scratch every time. It’s great for getting inspiration when I have an idea in mind but don’t quite know how to bring it to life. Review collected by and hosted on G2.com.

Sometimes it takes a few tries to get exactly what I have in mind, especially when it comes to small details or getting text right in an image. For example, I might love the overall design but notice that some elements are in the wrong place or that the image doesn't quite match my original idea. It can also be a bit frustrating when I want to make one small change and end up having to regenerate several versions. The results are great, but I wish it gave me a bit more control over the final output. Review collected by and hosted on G2.com.

Midjourney, hands down, delivers the best overall visual quality on the market. What I appreciate most is how naturally the images are rendered. Unlike other AI image generators where you’re constantly fighting the prompt, it’s much easier to get rich, dynamic results that don’t feel overly sterile or resemble the typical “AI stock photo” look. Review collected by and hosted on G2.com.

In my experience, Midjourney’s biggest drawback is how hard it can be to get precise, exact control over the final output. It can produce stunning imagery in a general sense, but fine-tuning specific details—such as exact object placement, accurate text rendering, or keeping identical character features consistent across more complex scenes—often ends up taking a lot of trial and error. Review collected by and hosted on G2.com.