Imagine a world where AI doesn't just mimic reality—it blurs the lines so seamlessly that even experts struggle to distinguish fact from fabrication. That's the thrilling yet unsettling frontier we're navigating in 2024's generative AI tools for images and videos, and marketers are buzzing with opinions on which ones truly deliver. Stick around, because this year's rankings reveal surprises that could change how you think about creativity in marketing—forever.
Generative AI tools have made leaps forward this year, edging ever closer to that eerie 'uncanny valley'—the point where AI creations look almost human but still feel just a tad off, sparking unease. Innovations from companies like OpenAI's Sora and Google's Nano Banana have inundated the market, leaving marketers with sky-high hopes for efficiency and cost savings. The initial excitement around these tools has faded, replaced by a demand for dependable, user-friendly options that slot effortlessly into existing creative workflows.
But here's where it gets controversial: Is AI really streamlining work, or is it complicating it by requiring constant juggling of multiple platforms?
'As Freddie Dabaghi, chief transformation officer at Crispin, wisely pointed out: 'If AI automates a task, but you end up shuttling files between 15 different software programs, you're not gaining efficiency—you're just complicating your day.'
This year, the AI landscape has produced clear winners, underperformers, and solid middle-ground options, evaluated mainly on how well they integrate, their ease of use, and the quality of their outputs. To shed light on their performance, Digiday compiled a 2025 agency generative AI report card, grading them on output quality, user experience, and overall capabilities. Keep in mind, this isn't a complete list, as new tools keep emerging rapidly.
For insights, Digiday contacted Google, OpenAI, and Midjourney. Google and OpenAI praised the artistic possibilities of their tools, while Midjourney didn't reply before our deadline.
Nano Banana: A Grade
True to Google's reputation for setting industry standards, its generative AI offerings are often seen as the benchmark by marketers. In November, Google unveiled Nano Banana Pro, an image generation and editing tool powered by its Gemini LLM.
Multiple agency leaders hail it as the ideal launchpad for creative projects. They praise its superior accuracy compared to rivals like Midjourney, Canva, or Adobe Firefly. For beginners wondering what makes this special, Nano Banana reduces what's known as 'AI sheen'—those jarring flaws, like unnatural perfection or inconsistencies that scream 'fake.' Instead, it delivers uniform, lifelike images that fool even the keenest eye, making it hard to spot AI involvement.
A major perk is its deep integration into Google's ecosystem, embedding directly into Google Ads, Gemini, and other workspaces for smooth campaign launches.
And this is the part most people miss: One candid creative director from a major global digital marketing agency (part of a larger holding company) called Nano Banana 'a pretty dang good image generation model,' adding it's their top choice for reliable work.
That said, the drawbacks stem less from Nano Banana itself and more from broader issues in generative AI. Marketers yearn for better uniformity across tools, ensuring storyboards, characters, and narratives remain consistent through iterations—something that's notoriously tricky, as a single adjustment can disrupt an entire AI-driven project.
VEO: A Grade
Leading the pack is VEO, Google's text-to-video AI, which has earned top marks for its versatility.
'VEO is the tool we rely on most—it's incredibly robust,' shared an anonymous executive from a branded entertainment firm. 'It's just pretty damn good at everything it does.'
Launched late in 2024, VEO gained momentum with the release of audio features in VEO 3.1 in May.
Like Nano Banana, VEO stands out in marketers' arsenals for its reliable output, including consistent character development, prompt interpretation, and cinematic execution. The addition of voice, sound, and lip-sync technology has elevated it further, some say even surpassing Sora.
Yet, despite these strengths, generative AI tools like VEO are still maturing. While Google's flagship products impress, VEO hasn't hit 'production-ready' status—free from that lingering AI sheen and polished for professional standards.
Sora: B Grade
OpenAI's Sora, a text-to-video model, shook up the visual AI world with its December 2024 debut. Its Sora 2 update, released in invite-only iOS beta in September, added authentic sound effects. Recently, a partnership with Disney let users incorporate Star Wars, Pixar, and Marvel characters, fueling excitement.
Marketers appreciate Sora's steadiness in depicting characters and settings, but the AI sheen remains a sticking point.
'Sora captures a lovely cinematic vibe, with excellent lighting and atmosphere, but it struggles in the uncanny valley—especially with skin textures,' noted a background agency group creative director.
While the Disney tie-in grabbed headlines, Sora has lagged behind competitors like VEO in realism and workflow integration. Plus, concerns persist about its training data, copyright issues, and intellectual property handling—topics that could spark heated debates among creators and rights holders.
Midjourney: B Grade
As a pioneer in AI visuals since 2022, Midjourney has evolved with updates for more lifelike images and user control. However, lawsuits from Disney and Universal have made some marketers hesitant. It trails rivals in hyperrealism, integrations, and interface usability.
One background creative director described it as ideal for 'Marvel-style movie shots,' but unreliable and risky for client projects due to data sourcing.
'It needs more precision,' said the agency group creative director. 'Midjourney feels like a roulette wheel—input a prompt, and you get four wildly varied outcomes.'
Compared to Google's Nano Banana or Adobe Firefly, it's often seen as outdated. At least one agency labels it a 'strict no' for professional use.
Conclusion: No Clear Champion Yet
A key takeaway—and one marketers emphasize—is that these tools work best in combination, creating a collaborative ecosystem. There's no undisputed leader, at least for now.
Overall, users criticize the lack of uniformity, control, and seamless connections. The goal of AI is to speed up, customize, and expand content production, but that vision remains aspirational amid a crowded, disjointed market.
But here's where it gets controversial: Are these tools empowering creativity, or do they raise ethical red flags about data privacy and job displacement?
'As Jon Weidman, SVP of development at Wavelength, a production company, put it: 'Consistency is the top priority for users. Until that's nailed, every platform must race to fix it.''
What do you think? Do you agree that Google's tools are reigning supreme, or is there untapped potential in underdogs like Sora? Is the push for hyperrealism worth the privacy risks from murky training data? Share your views in the comments—do these rankings change your perspective on AI in marketing, or do you see a different 'king' emerging?
