IMPLICATOR
Magnific, the Málaga company known until April as the stock-image site Freepik, renamed itself on April 28 and now sells subscription access to more than 30 outside AI image and video models through one platform. CEO Joaquín Cuenca Abela told Fortune the business reached $230 million in annual recurring revenue, with video generation accounting for roughly half, and that he built it without venture capital. His wager, as he described it to Fortune, is that creative teams will pay for one account that runs a wide range of leading models rather than licensing each one separately.
What It Does
- Magnific (formerly Freepik) sells one subscription spanning 30+ AI image and video models, upscaling and a 200M-asset library.
- CEO Joaquín Cuenca says it reached $230 million in revenue with no venture capital; private-equity firm EQT has held a majority stake since 2020.
- Named users include the BBC, Puma, Carl's Jr and Amazon Prime Video; pricing runs $10 to $280 a month plus team and enterprise tiers.
- The catch: AI output carries no copyright guarantee, plan credits expire each cycle, and the model catalog depends on outside providers.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
The company behind it
Cuenca, 49, cofounded Freepik in 2010 with brothers Alejandro and Pablo Blanes and grew it into one of the web's largest stock-image libraries before generative models threatened that business. He pivoted after OpenAI released DALL-E 2 in 2022, and in May 2024 Freepik bought Magnific, a Murcia upscaling startup founded by Javi López and Emilio Nicolás that had drawn more than 30,000 waiting-list signups in its first 24 hours. The April rebrand applied that acquired name to the whole company.
The numbers are unusual for a firm that never raised venture capital, though Magnific is not founder-controlled. The private-equity group EQT took a majority stake in 2020, with Cuenca and management keeping a minority. Fortune reported the $230 million revenue figure on April 28; Tech.eu put it at $200 million the same day. The company says it has more than one million paying subscribers and that its Business plan, launched in January, has been adding teams at roughly 150 a week.
Named customers run from the BBC and Guess to ad campaigns for Puma and Carl's Jr and the Amazon Prime Video series House of David, with Magnific citing between 250 and 290 enterprise teams depending on the announcement. "It's not that we are getting users from any particular competitor," Cuenca told Fortune. "It's that people are finding that they can do a new thing that was not possible before."
Where the risk sits
Because Magnific offers access to models from Google, OpenAI, Runway and ByteDance rather than training its own frontier system, its catalog depends on suppliers that can change prices or access. Its own documentation says it cannot guarantee copyright ownership of AI output in every jurisdiction and places infringement responsibility on the user. Cuenca has defended the underlying training practice in stark terms, telling Decrypt that requiring permission from every creator would mean the models "simply couldn't exist," likening it to asking permission to index every web page before launching Google.
Fortune reported that headcount fell to about 400 from roughly 550 in the stock-image era, with Cuenca conceding "there's going to be some pain in any transformation." On the App Store the product holds a 4.7-star average across about 1,600 ratings, though recent reviewers report paid subscriptions that fail to activate and support that has not answered them. Magnific's plan credits also expire each billing cycle rather than rolling over.
How it works
- Runs in the browser and in iOS and Android apps; an existing Freepik login carries over after the April rebrand.
- Search a library of 200M+ stock photos, vectors, icons, templates and video, or start from an uploaded or generated asset.
- Generate images from 30+ models, including Mystic, Google Imagen, Nano Banana, Flux, GPT, Ideogram and Seedream.
- Generate video from 30+ models, including Google Veo 3.1, ByteDance Seedance, Kling, OpenAI Sora, Runway and MiniMax Hailuo.
- Upscale and enhance images with the original Magnific technology, plus background removal, relight, restyle and expand.
- Build repeatable team workflows on Spaces, a node-based canvas with roles, comments and live cursors.
- Pick a model per task, let an
Automode choose one, or call everything through the API with anx-magnific-api-keyheader.
How to use it well
- Set a monthly credit cap in the dashboard before a big project. Video and 4K generations burn credits fast, and unused plan credits do not roll over.
- Lock composition in a cheap image model (
ClassicorFlux Fast), then upscale or switch to a premium model only for finals. - Use reference images and trained styles, characters and products to keep a face or product consistent instead of re-prompting from scratch.
- Build the campaign once in a Space and reuse the node graph for each new format or language rather than repeating steps by hand.
- For client work, confirm the license tier, and on Enterprise the indemnification terms, before shipping AI output.
- Avoid two reflexes: defaulting to the most expensive model for every generation, and treating AI output as copyright-clean stock, which Magnific's terms explicitly disclaim.
How it compares
The Next Web described Magnific as model-agnostic, competing with Midjourney, Runway and Adobe Firefly without matching any single one head to head.
| Tool | Paid floor | Differentiator vs Magnific | Native video | Stock library |
|---|---|---|---|---|
| Magnific | $10/mo | Many third-party models plus assets in one account | 30+ models | 200M+ assets |
| Midjourney | $10/mo | Image-generation specialist, no integrated stock library | Own model | No |
| Runway | $15/mo | Professional AI video generation and editing | Own Gen models | No |
| Adobe Firefly | $9.99/mo | Output designed to be commercial-safe, Creative Cloud-native | Own model | Adobe Stock (separate) |
| Canva | $15/mo | Template-driven design for non-designers | Limited | Elements library |
Get Implicator.ai in your inbox
Strategic AI news from San Francisco. No hype, no "AI will change everything" throat clearing. Just what moved, who won, and why it matters. Daily at 6am PST.
No spam. Unsubscribe anytime.
What it costs
The Business tier, aimed at teams, launched in January 2026; the rest of the range runs from a free trial to a $280-a-month Pro plan, all metered in credits.
| Tier | Price | What you get |
|---|---|---|
| Essential | $10/mo · $7.50 yr | 96,000 credits/yr, all image, video and audio models, commercial AI license, 200M+ assets |
| Premium | $20/mo · $14.50 yr | 240,000 credits/yr, more parallel generations |
| Premium+ | $45/mo · $33.75 yr | 600,000 credits/yr, credit-free generation on selected models, buy extra-credit packs |
| Pro | $280/mo · $210 yr | 4,000,000 credits/yr, credit-free on selected models; premium image, video and audio still consume credits |
| Business | $69/seat · $55 yr | 1.08M team credits/yr, 2 users, shared spaces, 8 parallel generations/user, API, basic single sign-on (SSO) and indemnification, 1-year rollover |
| Enterprise | Custom | Unlimited seats, custom credit allocation, custom SSO, SOC 2 Type I and ISO/IEC 27001, contract-based indemnification, admin controls, 3-year rollover |
The verdict
For a marketing team already paying for Midjourney, Runway and a separate stock subscription, Magnific's single bill and shared credit pool is a genuine consolidation, and a reported revenue figure that Fortune put at $230 million and Tech.eu at $200 million suggests enough buyers agree. The platform also draws on outside models it does not own, so its catalog follows those providers' prices and access, and its own terms decline to guarantee copyright on AI output. Whether the €10 million Magnific Fund, launched May 22 to help EU and UK creative teams adopt its Business plan through discounts and training, converts pilots into production contracts is the next thing to watch.
Pros
- One subscription spans 30+ image and video models, upscaling, audio and a 200M-asset library.
- Shared team credits, Spaces workflows and an API fit agencies and in-house teams.
- No venture funding, profitable by the CEO's account, and growing, with named enterprise customers.
Cons
- AI output carries no copyright guarantee, and infringement risk sits with the user.
- The model catalog depends on outside providers' prices and access.
- Plan credits expire each cycle, and video and 4K generations burn them quickly.
- App-store reviews flag billing and support problems.
Frequently Asked Questions
What is Magnific?
Magnific is the April 2026 rebrand of Freepik, the Málaga-based creative platform. It bundles AI image and video generation across more than 30 models, image upscaling, audio, a 200-million-plus stock-asset library and team workspaces called Spaces into one subscription, usable in the browser, on iOS and Android, and through a paid API.
How much does Magnific cost?
Individual plans run from $10 a month (Essential) to $280 a month (Pro), cheaper billed annually and each metered in credits. The Business team tier is $69 per seat per month, and Enterprise is custom and sales-managed. A free trial is available, and unused plan credits do not roll over to the next cycle.
Is there a free tier?
Magnific offers trial credits to start, and the older Freepik free library stays accessible. But AI generation, upscaling and video are credit-metered, so serious use needs a paid plan. Plan credits reset each billing cycle and do not carry over; only separately purchased extra credits last longer.
How does Magnific compare to Midjourney or Runway?
Midjourney and Runway each specialize, Midjourney in image generation and Runway in AI video. Magnific's pitch is breadth: many third-party models for both image and video, plus upscaling and a stock library on one bill. The Next Web has described it as model-agnostic rather than the best at any single task.
Who owns the copyright to what I make?
Magnific's own terms say it cannot guarantee copyright ownership of AI-generated output, which varies by jurisdiction, and the user is responsible for avoiding infringement. Paid plans include a commercial AI license, but AI output sits outside standard legal protection; Enterprise contracts may add indemnification.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Marcus Schuler
Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: [email protected]
