Generative AI tools are approaching the uncanny valley this year, with tools like OpenAI's Sora and Google's Nano Banana flooding the market. Now more than ever, marketers expect AI to streamline marketing processes and save money. As the novelty of generative AI wears off, marketers are looking for controllable and consistent tools that can be easily integrated into their creative ecosystems.
“If you're using AI to automate certain tasks, but you have to move files from 15 different pieces of software, you're not actually making your life more efficient,” said Freddy Dabaghi, chief transformation officer at Crispin.
This year, we have winners, losers, and in-between candidates based on integration features, user interface, and of course, final output. To better understand how these tools performed this year, Digiday has compiled a ranking of 2025 government-generated AI report cards based on outcomes, user experience, and functionality. This list is by no means exhaustive, as generative AI tools continue to flood the market.
Digiday has reached out to Google, OpenAI, and Midjoury for comment. Google and OpenAI defended the creative potential of their respective generative AI tools, but Midjourney did not respond by press time.
Nano banana: A
Like most things at Google, the company's generative AI tools are the gold standard, marketers say. Google brought Nano Banana Pro, an image generation and editing tool built on Gemini LLM, to market in November.
At least two agency executives said Nano Banana Pro is their go-to starting point for creative workflows, and raved about the image generation tool's higher accuracy than competitors like Midjourney, Canva, and Adobe Firefly. TThere's less of what marketers call AI gloss (described as inconsistencies or “too perfect” quality that people notice) and more consistency in what the image generation tools produce. The images look so real that it becomes more difficult for users to tell the difference between the real thing and the AI-generated one.
Another advantage of Nano Banana is that it is within the Google ecosystem. The image generation tool is embedded in Google Ads, Gemini, and other Google Workspaces, making it easy to go live in your campaigns.
“Nano Banana is a very good image-generating model,” says the agency's creative director, who agreed to speak candidly about the background. He added that Nano Banana is the agency's “favorite image generation tool.” The agency is a global digital marketing giant and part of a holding company.
Nano Banana's shortcomings are less about the Google-powered app itself and more about the overall standing of generative AI tools. Marketers want consistency across AI toolssThe ability to create storyboards so that characters, narratives, etc. look exactly the same across multiple iterations. This has been a struggle for marketers, as a small tweak to AI-powered creative can be a game-changer.
VEO:A
Also at the top of the class is VEO, Google's text-to-video generative AI model.
“VEO is probably the one we use the most because it's incredibly robust,” said a second anonymous source, an executive at a branded entertainment company, who spoke on condition of anonymity. He added: “Very good at everything.”
VEO was originally released in late 2024, but gained momentum in May with the rollout of audio features in VEO 3.1.
Like Nano Banana, VEO is another key tool in the generative AI toolkit, according to marketers Digiday spoke to for this article. VEO consistency in the ability to create characters, interpret and execute prompts and film concepts. Recently launched voice, audio, and lip-sync features make it an improvement over Sora, marketing officials say.
But again, VEO (and generative AI image tools, for that matter) is in its infancy. Marketers say Google's flagship product is good, but VEO has not yet reached production level, one that is polished, complete, and without any of the so-called AI sheen.
Sora: B
Sora, OpenAI's text-to-video AI model, revolutionized the landscape of image generation tools in December 2024. Its big step forward, Sora 2, was released on an invite-only basis to iOS users this September and features realistic sound effects. The turning point came earlier this month, when Disney and OpenAI entered into a partnership that allowed Sora users to use characters from Star Wars, Pixar, and Marvel series in their content creation.
According to marketers, consistency in characters, environments, and other aspects of image creation is the key to victory. But its main criticism is the brilliance of the AI.
“Sora had a really great cinematic quality to it. I think the idea of lighting and cinematic look is great, but there's a real uncanny valley in things like skin textures,” said a creative director at an agency group.
Although the Disney partnership made headlines, marketers say Sora's effectiveness lags far behind products like VEO, which are more surreal and easily integrated into other creative workflows. Also, business concerns regarding the use of Sora's training data, copyright, and IP persist.
During the journey: B
Midjourney can be considered the OG of the generative AI visual creation space. Since its launch in 2022, this image generation tool has steadily released updates that provide more realistic images and user controls. The tool has faced lawsuits from the likes of Disney and Universal, giving some marketers pause, but marketers say the tool leapfrogs competitors in terms of hyperrealism, integration and user experience.
Midjourney's is a more established brand perhaps better suited for “Marvel movie-style imagery,” but training data meant it was considered unreliable and commercially unsafe for client work, according to the creative director who first shared the background.
“Midjourney could benefit from more control,” said a creative director at an agency group. “Midjourney is a little more of that western slot machine. You type in a prompt and you get four very different results.”
After all, its place in the generative AI creative toolkit pales in comparison to Google's Nano Banana and Adobe Firefly. At least one agency called midjourney a “no-no” for client work.
Conclusion: There is no king to rule them all
What's important to note, and what marketers emphasize, is that many of these tools are used in conjunction with each other. There's no king to rule them all — at least not yet, marketers say.
Overall, marketers criticize its consistency, management, and integration. The promise of generative AI content creation tools is to speed up, personalize, and scale content creation. Marketers say that promise remains elusive because the onslaught of tools has fragmented the market.
“Consistency is probably the most important thing you have to offer people,” said Jon Weidman, senior vice president of development at Wavelength, an entertainment production company. He added: “Until that challenge is resolved, all platforms need to work toward solving the challenge.”
