Higgsfield, an image and video generation platform with roots in Kazakhstan, says it has surpassed a $1 billion annualized revenue run rate. Bloomberg reported the milestone on September 24, and CEO and co-founder Alex Mashrabov confirmed it on LinkedIn. According to the company, the business has grown twentyfold in a year.
On September 28, Mashrabov discussed the path to that result on Harry Stebbings’s 20VC podcast. Before its successful launch, Higgsfield spent a substantial share of its initial funding searching for the right product. The company now spends more than $4 million a month on its team’s internal use of AI models.
The sales milestone and the CEO’s account reveal two sides of the business: demand for AI-generated advertising content and the cost of building the platform that produces it. The reported billion dollars represents the current sales pace extrapolated over a year, rather than revenue already earned over twelve months.
What Higgsfield’s billion-dollar figure means
The company calculates its annualized revenue run rate by multiplying revenue from the most recent four weeks by thirteen. This estimates what the business could earn over 52 weeks if its current pace continued. The metric captures recent momentum but does not replace annual financial statements.
At a run rate of exactly $1 billion, that calculation would imply roughly $76.9 million in sales over four weeks. This is an arithmetic illustration, not a separate figure disclosed by the company. Actual results in subsequent periods will depend on new orders, recurring payments and customer churn.
RAEM previously reported that Higgsfield had reached a $500 million annualized revenue run rate. In August, the company reported $700 million. It also announced a $400 million Series B round led by DST Global at a $5.4 billion valuation. These figures describe different metrics: sales, capital raised and the investment valuation of the business.
The company spent most of its seed funding before taking off
In the 20VC interview, Mashrabov acknowledged that Higgsfield spent more than a year searching for product-market fit. In his account, the team spent more than $10 million of its $16 million in initial funding. He attributed the setbacks to his own focus on hype and a compelling company narrative rather than product quality.
According to the CEO, the turning point came after conversations with eight creative directors. They all highlighted one problem: AI video lacked camera control. Professional video production required the ability to specify camera movement, angles and visual sequences, rather than merely hoping a generation would turn out well.
This account explains Higgsfield’s product direction. Advertising teams need a tool that lets them reproduce results and control scenes. Generation becomes part of a production workflow involving a product, brand requirements, a script and multiple ad variations for testing.
Who pays for the platform
Higgsfield now offers image and video generation tools, effects, scene controls and templates. Marketing Studio focuses on advertising content: product images, videos, posters and marketplace assets.
For businesses, the service makes it faster to prepare creative variations. Advertisers can change the visual approach and compare ad performance, then scale production of the formats that work. This creates demand for regular use of the platform rather than one-off generation.
According to Mashrabov, quoted by Forbes Kazakhstan citing Bloomberg, revenue from enterprise customer contracts has grown tenfold since June. Subscriptions previously accounted for more than 90% of revenue; their share is now slightly above 60%. The remaining revenue comes from usage-based payments.
In a September statement, the CEO reported more than 32 million creators on the platform and use of its products at 78% of Fortune 500 companies. These are Higgsfield’s own figures; they do not disclose the size of individual contracts or the extent of adoption within those organizations.
Where $4 million a month goes
In the interview, Mashrabov described monthly spending of more than $4 million on internal AI model use by a team of roughly 400 people. That averages more than $10,000 per employee. The amount covers the team’s own work with models; treating it as the entire cost of customer generations would be incorrect.
AI is used for development, experimentation and building internal tools. According to the CEO, creative staff have also begun programming with the help of models. In this operating model, access to AI becomes a distinct working budget alongside payroll and infrastructure.
However, a large model bill alone does not demonstrate productivity gains. Its economic value depends on the results: which tasks the team completes faster, which tools it builds, and how much additional revenue or resource savings that produces. The disclosed spending is insufficient to calculate Higgsfield’s profit without the other cost categories.
Kazakhstani roots and the economics of growth
Higgsfield’s co-founders include chief technology officer Yerzat Dulat. The company is developing an engineering center in Almaty. Its connection to Kazakhstan includes local specialists contributing to a product for the international market.
According to Mashrabov, Higgsfield has had a positive gross margin since the start of 2026. This means revenue exceeds the direct costs included in providing its services. The metric does not disclose net profit: the company also funds development, employees, marketing and internal model use.
Higgsfield’s story combines the discovery of a specific customer need with rapid sales growth. Camera control helped the company build a product customers wanted, while advertising budgets supported its expansion. The next test is whether it can retain customers and sustain its production economics as generation volumes rise. The reported run rate of more than $1 billion shows the scale of current demand; the durability of the business will become clearer in subsequent reporting periods.
