Magnific announces over $300 million per year

Magnific, the Spanish AI creation platform formerly known as Freepik, announced on Tuesday that it has achieved over $300 million in annual recurring revenue, without having raised funds since 2010. The company plans to invest $10 million in the Asia-Pacific region by early 2027. No external funding is scheduled, but its CEO does not rule out the possibility of seeking it next year.
Professional Clients and Legal Guarantees
More than one million paying subscribers and over 400 corporate clients: these are the figures claimed by Magnific. For Joaquín Cuenca, co-founder and CEO of the company, the major evolution over the past year lies in the profile of the clients, which now includes film studios, large agencies, and marketing services of brands. Spontaneous acquisition has been complemented by organized prospecting: about 45 employees out of 370 are involved in pre-sales, sales, demonstrations, or support.
Creative professionals also work with clients to train them. Cuenca believes that this support is lacking in most AI labs. He describes subscribers as a reservoir of future professional contracts, with some introducing Magnific into their companies after years of use.
The CEO also attributes part of the growth among studios to the legal protections offered, which he considers essential before production use. The corporate plans include indemnification against third-party claims regarding the intellectual property rights of generated content. The creations belong to the clients; neither their input data nor the resulting productions are used to train models.
However, this protection excludes bad faith, Cuenca clarifies: obtaining Mickey Mouse after expressly requesting it makes the client liable. He advocates for a content-focused approach, with detection of unintentional infringements and user warnings. To illustrate that any tool can be used to violate rights, he cites the example of two overlapping circles that could infringe on the Mastercard brand.
Hagia Executes a Campaign with Spaces
In Singapore, Hagia Labs and its creative partner Superson produced films for StarHub's Unlimited+ packages. The Hublings, green characters from the operator, soar through the skies above Asia before arriving in Japan. Hagia designed the characters as well as the settings in Spaces, the tool offered by Magnific.
Hagia estimates that the time from order to delivery takes two to four weeks, compared to two to four months for traditional filming. According to Cuenca, the traditional method would have required teams, travel, and filming at each location.
The CEO describes a phased adoption: teams first complete their usual tasks faster, then tackle productions that were previously out of budget. He claims that some brands are replacing their static images accompanied by text with spots of cinematic quality, while studios are accepting fantastic or historical projects that were previously too costly.
Le brief IA que les pros lisent chaque soir
Les 7 actus IA du jour, décryptées en 5 min. Gratuit.
Inclus dès l'inscription : notre sélection des meilleurs guides & comparatifs IA.
Choisissez votre rythme
Gratuit · Pas de spam · Désabonnement en 1 clic
Cuenca then presents automation as the next step. Once a production process is assembled in Spaces, employees or their agents in ChatGPT or Claude can trigger it via Magnific's MCP server.
Magnific Relies on Seedance from ByteDance
Magnific relies on Seedance, developed by ByteDance, Kling from Kuaishou, Wan from Alibaba, as well as MiniMax models. According to Cuenca, several Chinese models are among the best for video, and users choose them for their quality. Until last year, its suppliers were American, and the CEO had never visited China; he has since made several trips, including a 20-hour visit. The country joins the top 10 markets for the company for the first time this year.
According to Cuenca, the Shanghai office is intended to facilitate exchanges with the teams designing these models, accelerate their addition to the platform, and relay user feedback. Magnific does not plan to develop a fundamental model to compete with Google, ByteDance, or OpenAI: Cuenca believes this battle is lost for an independent company. In his comparison with a MacBook, the models play the role of the processor; Magnific handles the tools, workflows, and controls surrounding it.
The other office currently being opened in Singapore is set to serve as a regional commercial base, alongside teams in South Korea, Japan, and Thailand. Cuenca reports a 35% increase in adoption by companies in the region, which is why he prefers to recruit locally rather than transfer staff from Málaga. Magnific plans to have over 20 employees in the Asia-Pacific region by the end of 2026, then double that in 2027. Product and engineering will remain at the headquarters in Málaga; the CEO presents these locations as a long-term commitment in Asia.
Client-Funded Growth and a Growth Objective
When the name change occurred in April, annual recurring revenue was at $230 million. Based on published figures, Magnific positions itself as the fourth largest European company in generative AI for this metric, behind Lovable, ElevenLabs, and Mistral. Mistral raised €3 billion in September; at ElevenLabs, a share sale doubled its valuation to $22 billion.
Cuenca asserts that Magnific remains self-funded and that client revenues have covered all its activities, including the transition to AI. The arrival of EQT as the majority shareholder in 2020 was through a buyback of shares from partners, without any cash injection into the company. The CEO aims to maintain the discipline he associates with this funding: only developing products for which clients are willing to pay.
EBITDA is positive, Cuenca claims. However, profits are limited by the reinvestment of most revenues into computing capabilities, he explains. He estimates that the necessary expenses for requested generations already amount to hundreds of millions of dollars, as users demand immediate access to new, more powerful image or video models.
According to Cuenca, markets are demanding companies that are significantly larger than they previously required; an IPO is not under consideration. His goal is to exceed $1 billion in revenue before the end of next year, a threshold he believes can be reached through organic growth. He appreciates not being imposed a timeline.
Brief IA — L'actualité IA en français
L'essentiel de l'actualité de l'intelligence artificielle, décrypté et expliqué chaque jour.