Madiha Yaqoob
Higgsfield, a two-year-old startup building AI video and image generation tools, has raised $400 million in Series B financing at a $5.4 billion valuation. The company announced the round on August 17, 2026, and said its annualised revenue has reached approximately $700 million. The valuation is more than four times the $1.3 billion the company carried at its previous round.
Who Is Backing the Round
DST Global led the financing. New investors include Tribe Capital, Growth Equity at Goldman Sachs Alternatives, Smash Capital, Fifth Wall, Valor Capital, Intel Capital, Liberty Global Tech Ventures, Mirae Asset Capital, and NTT DOCOMO Ventures. Existing backers Accel, Menlo Ventures, AI Capital Partners, GFT Ventures, Capra Ventures, BAM Corner Point, and BroadLight Capital also participated, according to the company’s announcement.
Higgsfield says it now has more than 30 million users across 238 countries and territories, and that users generate more than 20 million pieces of content each month. The company was founded by Alex Mashrabov, a former Snap executive.
The Shift From Creators to Businesses
The more interesting number in this story is not the valuation. It is where the revenue comes from. Higgsfield initially found traction with individual creators making social media visuals. Mashrabov told the Financial Times that businesses now account for most of the company’s revenue, compared with less than a quarter in January 2026.
That is a fast reversal, and it explains the pricing of the round. Consumer creative tools tend to churn. Enterprise content production is a much larger and stickier market, because advertising, marketing, product, and training teams all need video on a recurring basis and currently pay agencies and production houses to make it.
What This Means for the Content Market
If AI video generation gets good enough and cheap enough for commercial use, the constraint on how much video a company produces stops being budget and starts being ideas. A brand that could previously afford four campaign videos a year may start producing forty variations for different audiences, channels, and languages.
That changes the work rather than eliminating it. Someone still has to decide what the video should say, direct the tool toward a usable result, review the output, and fit it into a campaign. Those are creative and strategic skills, not button-pressing skills, and they are not automated by a better model.
Why Freelancers Should Pay Attention
For freelancers selling video editing, motion graphics, or social content, the practical takeaway is straightforward. Clients will increasingly arrive already knowing these tools exist, and price expectations for basic production will fall. The defensible position moves upstream toward concept, scripting, brand consistency, and knowing which shots genuinely need a camera and a human.
It is also worth noting that investors are now willing to assign multibillion-dollar valuations to AI application companies rather than only to foundation model labs. Higgsfield does not train the largest models in the world. It built a product people pay for on top of models. That is a route that is open to far more builders than frontier research is.
The Open Questions
Rapid revenue growth in AI applications has a history of being partly driven by early enthusiasm rather than durable contracts. The things to watch are enterprise retention over the next several quarters, gross margins once inference costs are fully accounted for, and how the company handles rights and likeness issues as generated video becomes harder to distinguish from filmed footage. Regulators in the European Union have already begun requiring machine-readable marking of synthetic content, and that obligation will shape how these tools are built.
Sources: PR Newswire and SiliconANGLE.
