AI Video Platform Higgsfield Secures $400 Mn Funding at $5.4 Bn Valuation

AI video platform Higgsfield has secured $400 Mn in a Series B financing round that values the start-up at $5.4 Bn, marking an aggressive valuation re-rating just months after its previous funding. The investment highlights a rapid shift in the generative video sector, as commercial momentum moves from creator-led adoption to business and enterprise workflows.

DST Global led the round, with participation from high-profile institutional funds and corporate venture arms. New backers included 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. Singapore-based 1982 Ventures said in a statement that it had also participated in the round through its second fund.

Existing institutional investors, including Accel, Menlo Ventures, AI Capital Partners, GFT Ventures, Capra Ventures, BAM Corner Point, and BroadLight Capital, also participated.

A Breakneck Valuation Step-Up

The $5.4 Bn valuation is more than fourfold higher than in January, when the San Francisco-based start-up was valued at $1.3 Bn after an $80 Mn extension that brought its total Series A funding to $130 Mn.

The valuation surge reflects steep growth in the company’s financial metrics. Higgsfield reported that its annualised revenue run rate reached $700 Mn in August, up from $200 Mn in January. While run-rate measures project current business velocity rather than completed 12-month audited revenue, the figures imply Higgsfield’s annualised revenue run-rate multiple is roughly 7.7x, up from at least 6.5x in January.

Higgsfield Growth Trajectory (January vs August 2026)

Valuation: $1.3 Bn ───► $5.4 Bn (+315%)

Revenue Run Rate: $200 Mn ───► $700 Mn (+250%)

Total Users: 15 Mn ───► 30 Mn (+100%)

The High-Margin Enterprise Pivot

More critical to the investment thesis than raw growth is a fundamental shift in Higgsfield’s customer mix. Alex Mashrabov, co-founder and chief executive officer of Higgsfield, told the Financial Times that corporate clients now generate the majority of the company’s revenue, up from less than 25% in January.

The company’s platform now supports visual production for 390 Fortune 500 companies across sectors, including advertising, marketing, retail, broadcasting and financial services. Total users have doubled to 30 Mn in 238 countries and territories, with the US remaining its largest market.

The changing revenue mix alters the underlying economics of generative video. While consumer applications drive high engagement, video rendering demands intensive computing infrastructure. Enterprise clients creating large-scale marketing and media assets have a predictable path to recurring revenue that can offset high structural costs.

Navigating a Shifting AI Landscape

Higgsfield is positioning itself in the AI media application layer rather than relying on the underlying infrastructure. Its ecosystem integrates proprietary tech with external foundation models, letting professional editors move between different production tools.

The market remains fiercely competitive, featuring tech-giant initiatives like Google’s Veo, ByteDance’s Seedance and Kuaishou’s Kling, alongside specialised AI firms like Runway. However, the application layer has gained premium importance following structural pullbacks from early pioneers. OpenAI discontinued its Sora web and app experiences on April 26, 2026, and is scheduled to retire the Sora API fully on September 24, 2026.

As core model capabilities commoditise, commercial victory increasingly belongs to software that translates raw AI generation into repeatable, automated corporate workflows. Higgsfield has leaned into this trend with its ‘agentic’ software, which automates multi-scene video creation. Following the May debut of its Supercomputer product, users of these automated tools grew 42-fold in three months, generating more than 20 Mn content pieces monthly.

A New Horizon for 1982 Ventures

The deal marks a significant milestone and a notable departure for Singapore-based fund manager 1982 Ventures. Holding a capital markets services licence from the Monetary Authority of Singapore (MAS) for venture capital fund management, the firm originally built its reputation as a seed-stage specialist targeting ASEAN fintech companies. Its debut fund closed at over $20 Mn in 2022, and it announced plans in 2023 to launch a second fund targeting $50 Mn. It has not disclosed a final close for Fund II.

1982 Ventures Fund Profile

• Licence Status: Capital Markets Services licensee, authorised by MAS for venture capital fund management

• Core Focus: Early-stage fintech, enterprise software and AI

• Asset Evolution: Moving from pre-seed/seed towards global AI growth; a second fund targeting $50 Mn was announced in 2023, with no final close disclosed.

While 1982 Ventures has broadened its mandate to include global enterprise AI infrastructure and automated workflows, the cheque is a significant step up from its traditional pre-seed and seed portfolio. The VC said it invested an undisclosed amount through Fund II and gave investors in its network co-investment access to the round.

“Higgsfield is exactly the kind of company we look for: a generational business being built at the leading edge of AI, with real revenue and real enterprise adoption behind it,” said Herston Powers, founding managing partner at 1982 Ventures.

Operational Scaling

The newly secured capital will fund deep R&D, expand global infrastructure and drive aggressive talent acquisition. A primary allocation will go towards stabilising backend access. “Compute is scarce,” noted Mashrabov, saying that the capital injection would allow Higgsfield to secure large-scale, long-term computational reservations.

Mashrabov and chief technology officer Yerzat Dulat founded Higgsfield in 2023, with chief strategy officer Mahi de Silva later joining the founding team. Earlier, Mashrabov helped establish computer-vision start-up AI Factory and served as its chief operating officer. The start-up developed the technology behind Snapchat’s Cameos feature. Snap acquired it in late 2019 for $166 Mn and publicly confirmed the transaction in January 2020.

Higgsfield’s latest financing arrives at a fundamentally different operational inflexion point than its earlier, creator-led expansion phase. The company is no longer asking investors to buy into the raw user-growth metrics of consumer AI video. Instead, its valuation is increasingly tied to its ability to capture recurring spending from corporate clients willing to pay premium rates for automated visual production at scale.

However, this transition presents a demanding commercial test. A valuation that has quadrupled in less than eight months assumes Higgsfield can convert its volatile revenue run-rate growth into a durable enterprise moat. It must do so while underlying models commoditise, competitors converge on identical workflows and computing infrastructure remains a punishing capital expense.

For 1982 Ventures, the investment carries a parallel strategic risk. Having built its reputation by securing ultra-early access to ASEAN fintech start-ups, the investor is deploying its second vehicle much further up the capital curve by backing a global AI company already valued north of $5 Bn. The ultimate marker of success will be whether the fund manager can leverage its expanded enterprise mandate and larger capital pool to secure repeatable access to competitive global rounds well beyond its seed-stage territory.