From $166M Snap Exit to $50M AI Video Startup — Alex Mashrabov

What does a founder do after a $166M exit? Alex Mashrabov built Higgsfield AI — and raised $50M to prove that the real AI video opportunity isn't consumer apps, it's the brands quietly producing 10,000+ ad creatives a year.

Key Facts

• Snap acquisition price: $166M for AR filter technology company (Alex Mashrabov)
• Higgsfield Series A raise: $50M, preceded by $8M seed led by Menlo Ventures (Alex Mashrabov)
• Top advertiser ad volume: Companies spending $10M+ on Meta launch 10,000+ new ad creatives per year (Alex Mashrabov)
• User engagement: Average Higgsfield users spend over 10 hours per month on the platform (Alex Mashrabov)
• AI video production cost: In-house AI-generated social videos can reach under $100 per video (Alex Mashrabov)

The Snap Exit and What Alex Learned About AI Video

Alex built the core AR filter technology that powered Snapchat's AI-generated face filters — enabling pixel-level AI generation on mobile for the first time. After selling for $166M and spending four years at Snap, he identified video as the underdeveloped frontier in generative AI.

Before Higgsfield, Alex spent a decade in video AI. His previous company built real-time face filter technology that ran neural nets on mobile phones — a capability that didn't exist before. 'We enabled these net new capabilities where every pixel is generated with AI,' he explained. 'ByteDance and Instagram didn't have those.' The acquisition contributed meaningfully to Snap's daily active user growth.

Working directly under Bobby Murphy and alongside Evan Spiegel gave Alex a front-row seat to product innovation at scale. 'Evan is probably one of the best product designers in the world,' he said. 'He's definitely the best product designer I have ever worked with.' That experience shaped his conviction that video — not images, not text — was the next major generative AI battleground. When he left Snap in September 2023, he moved immediately to found Higgsfield.

"I simply believe that there are going to be many more stories told using generative AI in video formats." — Alex Mashrabov

"It felt to me that video AI is a completely new space where we can jump in very early and ride the wave." — Alex Mashrabov

Why Consumer AI Apps Are a Trap — and B2B Advertising Is the Prize

Consumer AI apps attract users but rarely convert them to paying customers. Free-to-paid ratios hover around 5-6% — similar to traditional B2C SaaS — while mobile usage patterns reduce stickiness. The real monetization opportunity is B2B: professional marketing teams with real budgets and urgent production needs.

Higgsfield's early product chased consumer adoption. Features that let users make friends hug or kiss in AI-generated photos drove downloads but not revenue. 'There is no real business in that,' Alex said bluntly. OpenAI's willingness to subsidize consumer regenerations made that market structurally unwinnable for a startup.

The pivot insight came from observing CapCut. Originally a mobile-first TikTok editing companion, CapCut started appearing on desktops among professional social media teams. That signal pointed to an unmet need: marketing teams needed serious collaboration, approval workflows, and production scale — none of which mobile apps support. 'Most of the work in SMB marketing teams still requires some rounds of approval and collaboration, which is simply impossible to achieve on mobile,' Alex noted.

This led to the launch of Pixel.ai, Higgsfield's web-first platform. The target customer shifted from individual creators to professional marketing teams and agencies. Nine of the top ten major advertising agencies now use Higgsfield — not exclusively, but as a core tool alongside competitors like Runway.

"OpenAI clearly wants to take this market. They are willing to subsidize regenerations. That is why I am actually very happy that we decided not to pursue this anymore." — Alex Mashrabov

"It's very difficult to change the economics. When I look at free to paid ratio for OpenAI, we still talk about five, six percent. It's very similar to any B2C SaaS." — Alex Mashrabov

• Consumer AI free-to-paid conversion: ~5-6%, matching traditional B2C SaaS
• Mobile apps lack collaboration and approval workflows marketing teams require
• OpenAI actively subsidizes consumer regenerations, making it unwinnable for startups
• Web-first platforms unlock stickier professional use cases with higher willingness to pay

The Real AI Video Use Case: 10,000 Ad Creatives a Year

Top Meta advertisers spending $10M+ annually produce over 10,000 new ad creatives per year — roughly 40 per working day. Most teams cannot hit that volume manually. Higgsfield targets this gap by enabling AI-native marketing teams to produce, test, and iterate video ads at scale without physical production.

The numbers are staggering. According to Alex, large brand advertisers on Meta aren't producing a few hundred videos a year — they're producing tens of thousands. 'We are really talking about a very high pace. We are not talking about just four or five videos a day. We are talking about forty, at least forty videos a day, and this requires some level of automation.'

Higgsfield's sweet spot is 20-30 second videos designed for TikTok and Reels. Brands use the platform to create AI spokespersons of various demographics, run A/B tests on actor appearance, and iterate creative variables at a pace physical production can't match. The MrBeast model — obsessive A/B testing of thumbnails, opening seconds, and visual elements — is becoming the standard playbook for serious advertisers.

Cost benchmarks matter here. Hiring a Southeast Asia agency for AI-generated video runs $3,000-$5,000 per video. US agencies start at $30,000. But in-house teams using Higgsfield can produce quality social videos for under $100 each — comparable to what brands already pay TikTok creators for sponsored content.

"Companies which spend over ten million dollars in advertising on Meta are launching more than ten thousand new ad creatives every year." — Alex Mashrabov

"HiggsField empowers data-driven experimentation. They want to AB test how the actor should look like in the video." — Alex Mashrabov

Why Adobe Is in Trouble — and Where Higgsfield Fits

Adobe dominates professional image editing with Photoshop but has weak coverage of prosumer video creators. Premiere serves large production teams well, but the massive middle segment — independent creators and in-house marketing teams — has no dominant tool. Higgsfield is targeting exactly that gap as AI makes full-workflow video production possible in one platform.

Alex draws a sharp contrast between Adobe's image and video positions. Photoshop is genuinely universal — professionals and casual users alike. But in video, 'Adobe is very, very vulnerable.' Premiere serves large collaborative film teams well. Everything else falls short, leaving prosumers and small professional teams without a natural home.

CapCut fills the casual creator segment but lacks the professional workflow features marketing teams need. The analogy Alex reaches for is Figma versus Photoshop — a purpose-built challenger that captured a segment the incumbent couldn't serve. Higgsfield is positioning as that challenger for professional AI video.

The proof point is user behavior. Higgsfield users average over 10 hours per month on the platform. Image editing that once required exporting to Photoshop is now happening inside Higgsfield through prompt-based model interactions. 'I do not want to do clickbait for you,' Alex said, 'but it feels to me that Adobe is in huge trouble.'

"It feels to me that Adobe is in a huge trouble. People are making thousands of image editing operations on HiggsField on demand. It is an absolutely new economy." — Alex Mashrabov

"Adobe is very, very vulnerable, as Premiere has great penetration across professionals... but they don't really cover prosumers and individual professional creators." — Alex Mashrabov

How Higgsfield Found Product Market Fit — and What Signals Mattered

Product market fit came when professional creators and major brands started posting content made with Higgsfield organically — without any outreach from the team. The clearest signal was genuine organic reach from professionals, not just users acquired through the affiliate creator program.

Alex is candid that the first year was a struggle. The consumer pivot ate time and resources. But by January 2025, after going all-in on social media advertising, the signals changed. Organic creator posts about Higgsfield were outpacing paid placements five or six to one. Major industry figures were posting without being asked.

The go-to-market engine was built around AI educators — creators on YouTube, Instagram, and X who teach their audiences how to use AI tools. Higgsfield positioned itself as the best platform for product placement and brand-safe AI video, and those educators spread the word. 'Initially all the creators were posting for us for free,' Alex noted.

The product team operates on a relentless iteration cadence — daily updates except Sundays, with Saturday reserved for silent UI/UX improvements. That pace is intentional. 'The space is evolving so quickly. We see that major companies like ByteDance, Google, and OpenAI care about that space. We have to win by a faster pace of innovation and iteration.'

"We felt like we get true organic reach by professionals. It's not just our go-to-market efforts and creator program. It seems like the product is definitely resonating." — Alex Mashrabov

"The product has been live for more than two hundred days, and we still push updates, like product updates, every day except Sunday." — Alex Mashrabov

AI Video Production Cost Comparison