At Tech.eu, we talk a lot about scale, particularly how can small European startups grow into global players?
I recently caught up with Joaquín Cuenca, CEO of Freepik, at the company's Upscale Conf 2025 in San Francisco to explore just that.
Founded in Málaga, Spain, in 2010 by Alejandro Sánchez, Pablo Blanes, and Cuenca himself, long before the generative AI boom, Freepik has grown into the world's most visited generative AI creative suite, now boasting over 80 million monthly users.
Cracking the North American market is no small feat, yet Freepik stands out as one of the few European tech companies that has pulled it off.
I set out to learn how they did it, and along the way, I uncovered something resembling a playbook for today's ambitious European founders.
Beat the time zone
Cuenca attributes the company's success to intensity and speed.
"Being based in Europe is a disadvantage; we're not in the same time zone as the tech hubs, and sometimes we wake up to find two competitors have already launched something overnight."
So, from the get-go, Freepik worked harder. When it launched our image-generation API, we did it at 1 am to be first.
"That's the level of commitment we've maintained," contends Cuenca.
"On top of that, we built a local team in the US to do things we simply couldn't do from Spain — boots-on-the-ground marketing, partnerships, and presence at events. We weren't afraid to invest there."
Reinvent company culture for the GenAI era
After 10+ years, Freepik's culture has evolved. Cueno contends that around 2020, it had a very stable product-market fit.
"We knew our growth engine—better stock images, then illustrations, icons, 3D, templates, and so on.
It was predictable, and our culture adapted to that stability."
But GenAI changed everything. Suddenly, admits Cuenca, they didn't know what to build.
"We had to move fast, break things, and iterate. That meant going back to the fast-paced, experimental mindset we had in the early days. The culture shifted again to match that urgency."
Face the GenAI tsunami and ride the wave
Cuenca acknowledges that while many designers have made the leap to GenAI, others are hesitant.
"They're afraid — afraid of losing their jobs or seeing their years of hard-earned skills become obsolete. We get that. We've been there too."
He asserts that Freepik didn't start as an AI company.
"When DALL·E 2 launched just three years ago, it felt like a tsunami was heading straight toward us. I even got sick from the stress — I was feverish for a week, thinking all we'd built would be swept away.
But we soon realised that what felt like a threat was actually an opportunity, a wave we could ride. The key was remembering why we do what we do, not just how we do it."
Fake it until you make it
Freepik has formed some impressive partnerships. Cuenca admits that in the early days, Freepik "had to fake it a bit—by being so fast and independent with our tech that it looked like we were already working with these companies.
"For example, when CLIP from OpenAI was released, we integrated it in under an hour. We didn't wait for a partnership to form.
That kind of speed and visibility got us noticed.
Google reached out to us after noticing the high number of people using their models through our platform.
It's the same story with others. We move fast and deliver real business results. That makes us an attractive partner."
In February, Freepik integrated Google DeepMind's Veo 2 video model into its Premium+ AI Video Generator tool. This collaboration positioned Freepik as one of the first platforms to offer this technology, even before Google itself.
Work fast, ship faster
Six months ago, Freepik hosted its first conference in Málaga, Spain. Since then, we've launched dozens of features. It releases one major feature every 3.7 days on average.
According to Cuenca, the company's speed to ship features is cultural:
"We went back to our startup roots. We gave small teams complete ownership and removed red tape. We asked tough questions —like, "Why does this need to take three days? Can it be done by 3pm?"
Cuenca contends that often, delays come from layers of assumptions. For example, a tech lead adds buffer time, then the PM adds even more.
"No one really interrogates the estimate. But we do. Many of our engineering leads are former founders; they've done it themselves, and they know how fast it can be.
Cuenca believes that once you start challenging assumptions, people realise, "Oh, actually, this isn't hard." After doing that a few times, the mindset sticks.
Teams begin to estimate from first principles. They understand what needs to happen and how long each part really takes. That's how we maintain our speed."
Focus on cross-pollination in product design
Freepik sits at the intersection of tech and creativity. I wanted to understand how the company balances those two worlds.
While creatives and engineers do think differently, at Freepik, they work side-by-side. Cuenca shared:
"The tech teams build the tools, but they have to understand the creative context — what people are trying to express, how the tools are used.
We encourage deep collaboration. Engineers need to understand what's possible creatively, and creatives need to understand the limits and possibilities of the tech. That cross-pollination drives innovation."
Build your multimedia empire through acquisition
Since 2022, Freepik has made a series of strategic acquisitions to broaden its content offering, enhance its AI capabilities, and expand its global footprint. These moves have significantly expanded Freepik's media range — from static vectors and photos to video, audio, and 3D mockups — while also bolstering its technological edge and deepening its reach across Europe, Latin America, and North America.
In June 2022, the company acquired Videvo, a UK-based stock video and audio platform, marking its official entry into motion and sound.
This was followed by the October 2022 acquisitions of Iconfinder, a major Danish icon marketplace later integrated with Flaticon to create a global leader in icon design, and Original Mockups, a Colombian provider of 3D mockups and templates that helped expand Freepik's presence in Latin America. In October 2023, Freepik acquired EyeEm, a German photo marketplace with a library of around 160 million images—part of a broader push into generative AI and the US market.
Most recently, in May 2024, it acquired Magnific, a Spanish AI startup specialising in image upscaling and enhancement, marking its most AI-centric acquisition to date.
According to Cuenca, the company's largest untapped market is China.
"It's not just the language barrier — it's infrastructure. People there don't use email, Google, or Facebook, they use WeChat.
The whole user flow has to be reimagined. We've outlined six or seven things we'd need to localise to get traction in China."
Cuenca admits, "We see image generation becoming commoditised soon, so we're preparing to expand. We've always been looking for more companies to acquire."
Expand with ethical, legal products
The company recently launched Freepik Enterprise, a scalable, AI-powered creative suite designed for large organisations seeking flexible, secure, and legally compliant content generation.
Clients benefit from full ownership of generated content, legal indemnification, and strict privacy protections — Freepik guarantees that no enterprise data is used to train its AI models.
Cuenca detailed:
"Smaller businesses usually don't have their own models, but bigger ones like fashion brands often do. They want to integrate their models into our UI.
That's where we come in. We offer the infrastructure and UX for them to plug in their own models without having to build a front-end from scratch.".
In April, Freepik launched a new generative AI image model called F Lite, developed in partnership with AI startup Fal.ai and trained over two months using 64 Nvidia H100 GPUs. The model, which contains around 10 billion parameters, was trained exclusively on 80 million commercially licensed, safe-for-work images, making it part of a small group of image models built on licensed data — a key issue amid ongoing copyright lawsuits involving companies like OpenAI and Midjourney.
It recently partnered with CAPSI, a GenAI-driven studio based in Amsterdam and Barcelona, to create the opening and closing titles of the event.
The result: The Shape of Becoming, the story of Emmy, a woman who follows a mysterious black cube through surreal, AI-generated worlds. It's definitely worth a watch:
Cuenca asserts:
"What I loved most about this video is how it shows AI as a tool to speed up ideas, not replace them. It removes friction. It brings ideas to life faster. That's what excites me about GenAI."
Lead image: Freepik.
Would you like to write the first comment?
Login to post comments