Oct 13, 2025
Alex Armstrong
Which visual AI tools are creatives actually choosing for professional work, and why?
We reached out to over 8,000 creative professionals across the globe to understand the tools reshaping their workflows, the gaps still frustrating them, and where the industry is heading. The results reveal some clear trends in the viability of AI tools in professional creative settings.
ChatGPT's Narrow Lead In Image Generation
When we asked creators to rank their preferred AI image tools, one result stood out: ChatGPT emerged as the top-ranked tool with an average ranking of 2.66 out of 6. Midjourney follows closely behind with an average ranking of 3.06.
Why ChatGPT? The answer lies in accessibility and integration. Creators told us they value tools that fit seamlessly into their existing workflows. ChatGPT's combination of text understanding, image generation capabilities, and conversational interface makes it the most versatile creative tool. Especially for those who work across multiple mediums.
The picture gets more nuanced when we look at single-tool loyalty. When asked, “If you had to choose a single image AI tool to use for the next year, which would it be?” our respondents split nearly evenly between Midjourney and ChatGPT, with Midjourney edging ahead by just 5%. This tells us that there is still plenty of room in the market for specialized visual AI tools, particularly among professionals in established creative industries.
Video AI: The New Frontier
As for the video AI landscape, two names dominate the conversation. Veo 3 and Sora 2 captured over 50% of creator preference when asked to choose a single video AI tool, with Veo 3 taking the edge in this category by just 1%.
What makes these tools stand out? According to our respondents, it's about the quality of motion and adherence to prompts. While image AI has reached impressive photorealistic quality, video AI is still finding its footing. Creators gravitate toward tools that can maintain consistency across frames and actually deliver what they ask for, which is a capability that remains elusive for many platforms.
Midjourney and Kling AI battle for the second tier of preference, with a collective 34% of selections.
The Self-Employed Creator Economy
Perhaps the most telling insight from our survey isn't about tools at all—it's about who's using them. Self-employed creators represent the largest segment of visual AI adopters, comprising 34% of our respondents. This isn't surprising; these are the professionals who benefit most from tools that multiply their output without multiplying their costs.
The technology sector came in second at 13%, reflecting a still nascent adoption rate in this space. Media/entertainment firms and architecture and interior design round out the top adopters, each seeking to accelerate their creative needs. Based on the data, our hypothesis here is that these tools are not yet delivering quality standards that larger organizations demand. In professional applications where less photorealism and polish is required, visual AI tools are more often adopted.
Photography Dominates Project Types
When it comes to use cases, photography and photo manipulation lead by a significant margin, appearing in 40% of responses. This is unsurprising, as generating and editing images is where AI tools have reached their most mature state. Self-employed photographers and content creators are using AI to expand their portfolios, create variations, and explore creative directions that would be impractical with traditional photography.
The Frustration Gap: What's Still Missing
Despite the impressive capabilities of modern visual AI tools, creators are vocal about persistent frustrations. We categorized their feedback into four key pain points:
Fine Details and Quality emerged as the #1 concern, with creators frustrated by inconsistent textures, artifacts, and lack of precision in small elements. As one respondent put it: "Better sharpness and more detailed, realistic textures would help a lot."
Prompt Understanding came in close second. The challenge of translating creative vision into machine-readable instructions remains a significant barrier. Multiple creators mentioned struggling to describe complex scenes or finding that AI tools "improvise too much" rather than following specific instructions.
Consistency and Control ranked third, with creators emphasizing the difficulty of maintaining character consistency across images, achieving repeatable results, and exercising precise control over composition. One graphic designer from Georgia summed it up: "shorter prompts, better understanding."
The infamous "Hands and Text" problem—AI's notorious struggle with anatomically correct hands and readable text received substantial mentions. As one educator from Chile colorfully noted: "Finger and teeth like a scary tale."
The Implications: Where We Go From Here
Two major trends emerge from this data:
The consistency challenge
As AI-generated content moves from one-off experiments to professional workflows, the ability to maintain style, character, and quality consistency becomes paramount. The tool that overcomes this problem will be well positioned to capture significant market share.
The editing revolution
The next frontier is precise, localized editing. Creators want to modify parts of generated images and videos while preserving the rest, similar to traditional photo editing workflows. As it stands today, tools are falling short in these abilities.
Looking Ahead
The visual AI landscape is rapidly evolving. While ChatGPT, Midjourney, Veo 3, and Sora 2 have captured early mindshare, the rapid pace of development means leadership could shift quickly. The tools that win long-term will be those that solve the consistency problem, improve prompt understanding, and give creators the control they need to refine AI output into professional work.
Methodology: This analysis is based on survey responses from creative professionals across 40+ countries, representing diverse industries including self-employed creators, technology companies, education, media, and design agencies. Respondents ranked their tool preferences and provided detailed feedback on capabilities and pain points.