Higgsfield AI Raises 400 Million Dollars At 5.4 Billion Valuation As Enterprise Video Generation Demand Explodes
Higgsfield AI Raises 400 Million Dollars At 5.4 Billion Valuation As Enterprise Video Generation Demand Explodes
Highlights:
- Higgsfield AI closed a 400 million dollar round at a 5.4 billion dollar valuation
- The round was led by DST Global, Goldman Sachs, Liberty Global and Intel
- Valuation has more than quadrupled from 1.3 billion dollars just eight months ago
- Annualised revenue jumped from 20 million dollars to 700 million dollars in a year
- Enterprise customers now make up the majority of revenue, up from under 25 percent
- The platform reportedly generates about 4.5 million video clips every single day
Two years is not a long time in most industries, but in the current generative AI market, it can be long enough to build, scale and revalue a company several times over. Higgsfield AI, a generative video platform founded in 2023 by former Snap executive Alex Mashrabov alongside co-founder Yerzat Dulat, has just closed a funding round that puts a number on exactly how fast that compression has become, raising 400 million dollars at a post-money valuation of 5.4 billion dollars, a jump of more than four times its 1.3 billion dollar valuation from just eight months earlier.
The scale of investor appetite behind this round is itself a story worth sitting with. The financing was led by DST Global, the fund built by early Facebook backer Yuri Milner, alongside Goldman Sachs, Liberty Global and Intel, four names that do not typically chase speculative early-stage startups without something concrete to point to. Additional participation came from Tribe Capital, Smash Capital, Fifth Wall, NTT DOCOMO Ventures and Valor Capital, rounding out a syndicate that spans traditional investment banking, telecommunications infrastructure and venture capital simultaneously—a combination that reflects how broadly generative video has started to matter across industries rather than staying confined to the media and entertainment world it originally emerged from.
To understand why capital is moving this quickly, it helps to trace Higgsfield’s own trajectory rather than just look at the headline valuation. The company launched its browser-based generative video product in 2025, positioning itself around making sophisticated video generation tools accessible to creative professionals, marketers and social media creators, without requiring the technical overhead that earlier generative video tools demanded. It raised a 50 million dollar Series A in September 2025, valuing the company at around 1 billion dollars, then closed an 80 million dollar extension in January 2026 led by Accel at a 1.3 billion dollar valuation. From there, the growth curve steepened sharply. By June 2026, the company had crossed a 500 million dollar annualised revenue run rate, up from around 200 million dollars at the end of 2025, and by the time this latest round closed, annualised revenue had reportedly reached 700 million dollars, up from just 20 million dollars a year earlier, one of the fastest revenue ramps recorded in software history.
What has driven that ramp is a decisive shift in who is actually paying for the product. Higgsfield began, in large part, as a tool aimed at individual creators and entertainment-focused use cases, but its own disclosures show enterprise customers now account for the majority of its revenue, up sharply from less than 25 percent as recently as January 2026. Hundreds of direct-to-consumer brands, including Dollar Shave Club, are now reportedly using the platform to batch generate marketing videos on a daily basis, treating it less as a novelty creative tool and more as production infrastructure for advertising and social content at scale. That shift, from consumer curiosity to enterprise dependency, is precisely the kind of transition investors tend to reward most aggressively, because it suggests a revenue base built on recurring operational need rather than one-time experimentation.
“The shift from consumer curiosity to enterprise dependency is precisely the kind of transition investors tend to reward most aggressively. By building a cinematic logic layer directly into production workflows, Higgsfield has transformed generative video from an experimental novelty into essential enterprise infrastructure for global marketing operations.”
The competitive backdrop also helps explain the timing of this round. OpenAI’s Sora, once seen as the most prominent entrant in generative video, was shut down earlier this year after reportedly facing inference costs estimated at around 15 million dollars a day, a scale of expense that even a company backed by OpenAI’s resources found difficult to sustain against its usage patterns. Runway, long considered Higgsfield’s closest direct competitor in commercial video generation, closed its own large round, a 315 million dollar Series E at a 5.3 billion dollar valuation, back in February 2026, but has since signalled a strategic pivot away from commercial video production and toward what it calls world models, AI systems designed to simulate physical environments for applications in gaming, robotics and medicine rather than marketing content. That pivot has effectively left a meaningful portion of the professional video production market less contested than it was at the start of the year, and Higgsfield appears to be the company best positioned to absorb that vacated ground.
Numbers aside, what Higgsfield has actually built technically is also worth understanding on its own terms. The company describes what it calls a cinematic logic layer, a reasoning engine that sits between a user’s creative intent and whatever underlying generation model is doing the actual rendering. The idea is that when a user types an instruction like “make it dramatic,” that phrase is not something a raw video generation model can directly execute; it requires an intermediate layer of reasoning to translate loose creative language into specific, executable production choices—camera movement, lighting, pacing and framing among them. According to the company’s own figures, this pipeline currently produces around 4.5 million video clips every day, a volume that also points to substantial and continuously growing computing infrastructure demand, with the company confirmed to be running at scale on Nvidia hardware.
The new capital, according to the company, will be directed primarily toward enterprise-grade product development, security infrastructure and reserving additional compute capacity—three priorities that read less like a startup chasing growth for its own sake and more like a company preparing to defend an enterprise customer base that now expects reliability, data security and scale commitments typically associated with established software vendors rather than a two-year-old startup. That framing matters because enterprise buyers, particularly at the scale of consumer brands running daily marketing operations through the platform, tend to demand a different tier of operational maturity than early-adopter consumers ever did.
None of this is to suggest the trajectory is without real risk. A 5.4 billion dollar valuation against roughly 700 million dollars in annualised revenue implies a revenue multiple that, while not irrational for a software company still growing this quickly, does leave limited room for a slowdown before investor sentiment would likely reprice the business sharply downward. The broader generative AI video category has also drawn genuine public criticism—YouTube and other platforms have been actively cracking down on what has been termed “AI-generated content slop,” low-effort, high-volume synthetic video flooding platforms and frustrating human creators, a dynamic that creates real reputational risk for any company whose core product is, at its core, a very fast video generation engine. Higgsfield’s own bet is that enterprises will pay specifically for high-quality, brand-safe output rather than indiscriminate volume, a bet that so far appears to be working commercially, but one that remains dependent on the company continuing to out-execute both well-capitalised rivals and a steadily improving pool of foundation video models.
Taken as a whole, Higgsfield’s rise captures something distinct about where AI investment enthusiasm currently sits: less focused on raw foundation model scale and increasingly focused on companies that have found a genuine, monetisable enterprise wedge on top of that underlying technology. Whether a 5.4 billion dollar valuation proves justified will likely hinge less on this round’s headline number and more on whether Higgsfield can keep converting its explosive revenue growth into the kind of durable, defensible enterprise relationships that outlast the current wave of investor enthusiasm for anything labelled generative video.
