In a funding round that underscores the frothy investor appetite for generative AI tools, Higgsfield, a San Francisco-based startup specializing in AI-driven video creation, has secured an $80 million Series A extension, catapulting its valuation to more than $1.3 billion. Founded by Alex Mashrabov, a former Snap executive who led generative AI efforts at the social-media giant, the company claims a $200 million annualized revenue run rate just nine months after launch. This extension builds on an initial $50 million Series A, bringing total funding to $130 million.

Higgsfield’s rapid ascent comes amid explosive demand for AI video generation, where tools can produce high-quality clips from text prompts, transforming marketing, content creation and social media. The startup, which emerged from stealth in April 2025, now boasts 15 million users and partnerships with major brands. Investors including Accel led the latest tranche, drawn by metrics that rival established players in the sector.

From Snap to Startup Helm

Mashrabov, who spent over a decade at Snap Inc., rising to head of generative AI, spotted an opportunity in video synthesis during his tenure. At Snap, he oversaw features like AI-generated lenses and avatars, but constraints of a public company limited bolder experiments. ‘We were building for hundreds of millions of users, but now I can move faster,’ Mashrabov told TechCrunch.

Launching Higgsfield in early 2025, Mashrabov assembled a team of ex-Snap engineers and AI researchers from OpenAI and Google DeepMind. The platform’s core innovation lies in its diffusion models fine-tuned for cinematic video, enabling outputs up to 30 seconds long at 1080p resolution. Early adopters include ad agencies and TikTok creators, who praise its realism and customization.

Revenue Rocket Fuel

Higgsfield’s $200 million run rate stems from a freemium model: free basic generations hook users, while pro tiers at $20-$100 monthly unlock unlimited credits, 4K exports and brand kits. Enterprise deals, averaging $500,000 annually, target Fortune 500 marketers needing scalable video production. ‘We’re replacing entire creative teams,’ a company spokesperson said, per Reuters.

Competition intensifies from OpenAI’s Sora, Runway and Pika Labs, yet Higgsfield differentiates with ‘character consistency’—ensuring people and objects remain coherent across clips—and integration with Adobe Premiere. User growth hit 15 million by Q4 2025, fueled by viral X posts showcasing celebrity deepfakes and product demos.

Investor Frenzy in AI Video

Accel partner Margaret Francis highlighted Higgsfield’s traction: ‘Their revenue per user is 5x industry averages, with churn under 3%.’ The round valued the firm at $1.3 billion post-money, implying a 10x multiple on run-rate revenue. Other backers include Thrive Capital and former GitHub CEO Nat Friedman. Posts on X from TechCrunch amplified the news, garnering thousands of views and sparking debates on AI’s creative disruption.

This funding reflects broader VC momentum: AI video startups raised over $2 billion in 2025, per PitchBook data. Higgsfield’s path mirrors Mistral AI’s valuation leaps, betting on multimodal models as the next frontier post-text and images.

Technical Edge Under the Hood

At its core, Higgsfield employs a proprietary ‘Video Latent Diffusion’ architecture, trained on 100 million hours of licensed footage. Unlike competitors relying on public datasets prone to biases, Higgsfield partners with Hollywood studios for clean data. Mashrabov emphasized safety: ‘We watermark every output and block harmful prompts at inference time.’

The model’s efficiency—generating a 10-second clip in under 30 seconds on H100 GPUs—stems from custom quantization techniques. Upcoming releases promise audio sync and multi-shot narratives, positioning Higgsfield for Hollywood pilots already in discussion, as reported by BitcoinWorld.

Monetization Mastery

Enterprise wins include Coca-Cola and Nike, using Higgsfield for personalized ad variants. One campaign generated 10,000 unique 15-second spots, slashing costs by 80%. ‘Run rate hit $200 million organically—no marketing spend,’ Mashrabov noted in a Tech Startups interview.

Challenges persist: high compute costs burn $50 million yearly on Nvidia chips, offset by revenue but pressuring margins. Regulatory scrutiny looms, with EU AI Act classifying video gens as ‘high-risk.’

Path to Dominance

Looking ahead, Higgsfield eyes a Series B at $3 billion valuation by mid-2026, per sources close to the company. Acquisitions of smaller tools like Luma AI are rumored. Mashrabov envisions ‘AI as the new camera,’ democratizing filmmaking. As Forbes detailed, brands now view it as core infrastructure, producing thousands of assets weekly.

Risks include model commoditization as open-source alternatives emerge. Yet with 70% gross margins and 300% quarter-over-quarter growth, Higgsfield leads the pack. X sentiment buzzes positively, with creators hailing its ‘Photoshop for video’ leap.

Broader Market Ripples

Higgsfield’s success signals video AI’s shift from gimmick to business imperative. Traditional VFX firms like Industrial Light & Magic test integrations, fearing obsolescence. Investors pour in, with similar rounds at Runway ($300 million) and Kling AI ($1 billion).

For insiders, the real story is defensibility: Higgsfield’s moat combines data, distribution via Snap alumni networks, and relentless iteration. As Mashrabov put it to NewsBytes, ‘We’re not just building tools; we’re redefining creation.’