Where the money goes: AI fashion-tech funding in 2025 and 2026

On 22 October 2026, Alta, an app that builds a digital wardrobe and dresses an avatar of its user, announced a USD 20 million Series A led by Forerunner, which pre-empted the round. Its earlier backers, Menlo Ventures and Aglaé Ventures, the Arnault family investment firm, came back in. Sixteen months earlier Alta had raised an USD 11 million seed, also backed by Conviction and by angels including Karlie Kloss and the stylist Meredith Koop. According to the company, the app runs on more than a dozen in-house, fashion-specific generative AI models. Between the two rounds it signed Rent the Runway, whose customers, during a pilot of Alta's AI Outfits feature, added items to their bag 12% more often.

The round captures where fashion-tech money is going in 2025 and 2026: towards consumer-facing AI that styles, searches and shops on a customer's behalf, and towards tools that compress design and content work. It is going there against a weak backdrop. Crunchbase data show global e-commerce startup funding at around USD 7.3 billion in 2025, a fraction of the peak in 2021. The direct answer to where the money goes is that it is concentrated in a small number of AI-native companies with well-known founders or investors, while older fashion platforms are being sold at steep discounts and some AI startups have already closed.

For brands and retailers this matters because these startups are increasingly their partners, their distribution channels and, in some cases, their competitors for the customer relationship. Knowing which categories attract capital, and which fail, helps decide whom to work with.

How much venture money goes into AI fashion startups?

Less than the headlines suggest. In July 2025, Crunchbase News reported that funding to startups combining AI and apparel peaked at USD 162 million in 2022 and has averaged about USD 100 million a year since. Its list ranged from trend prediction and demand planning to design tools, virtual try-on, textile defect detection and recycling. Fashion is a small corner of AI investment: a single seed round, Daydream's USD 50 million in June 2024, amounted to about half of a typical year's total in the category.

The wider pools are much larger, and they are measured differently. Crunchbase, as reported by Fortune in November 2025, put global funding to e-commerce-related startups at around USD 7.3 billion in 2025, compared with more than USD 94 billion in 2021 and around USD 31 billion in 2020. CB Insights, in its State of Venture report published in January 2026, counted USD 27.5 billion invested in retail tech in 2025, up 7.4% on the previous year, across 2,088 deals, 20% fewer than a year earlier, with the average deal size up 34% to USD 20 million. The two trackers define the sector differently, so their totals cannot be compared, but both point to fewer, larger bets.

Indy Guha, general partner at VMG Partners, summarised the split for Fortune: "the destination side is down, enabling technology is up." New shopping destinations, he added, need "a really crisp reason to exist." That tension runs through the fashion deals of the past two years.

Which AI fashion startups raised money in 2025 and 2026?

The verified rounds fall into a few clusters. Consumer styling and shopping apps raised the most visible money. Design and content tools raised the most from the most prominent venture firms. Planning and inventory software raised steadily but quietly. The timeline below shows the main deals and failures that could be verified from primary or quality sources.

Daydream, founded by Julie Bornstein, a former executive at Nordstrom, Sephora and Stitch Fix, is the reference point for the consumer category. It raised its USD 50 million seed in June 2024 and opened its chat-based shopping assistant to all users in June 2025 with more than 8,000 brands, which it onboards free of charge in exchange for a percentage of each sale. Its technical bet is on understanding products in detail: Maria Belousova, who joined as chief technology officer in 2025, said the team was working to recognise "stylistic attributes like embellishments, silhouette" rather than relying on generic product descriptions. Bornstein told TechCrunch at launch that in e-commerce "search has always been the forgotten child, and it never worked very well".

Phia, co-founded by Phoebe Gates and Sophia Kianni, followed an USD 8 million first raise in September 2025 with USD 30 million in December at a USD 180 million valuation, according to eWeek. The app, which compares prices for items shoppers are looking at, reported 750,000 downloads in the eight months to November. Phia's round was led by Hans Tung of Notable Capital, with Kleiner Perkins and Khosla Ventures participating, and the app counts Hailey Bieber, Kris Jenner, Sheryl Sandberg and Sara Blakely among its supporters. Celebrity backing has become a feature of the consumer category, and it buys attention more than it proves a business model. Doji, founded by Dorian Dargan, who worked on visionOS at Apple, and Jim Winkens, a former DeepMind researcher, raised USD 14 million to build photorealistic avatars on which users can try clothes found anywhere on the web. Users take six selfies and upload two full-body photographs, and the app was invite-only across more than 80 countries at the time of the round. It could not yet show how a garment fits, a gap that matters for any retailer hoping such tools will reduce returns. "We saw Lensa and thought, what if we could do this for fashion, but in a photo-realistic way?" Dargan told TechCrunch.

Why are investors betting on AI stylists and shopping agents?

Because they promise to own the moment of choice. If a shopper asks an assistant what to wear to a summer wedding, whoever answers decides which brands are shown. Alta's case shows how such apps are trying to turn consumer reach into revenue from brands. Alta now powers the AI Outfits and Avatars features in Rent the Runway's app, where the feature reached 35% engagement and lifted add-to-bag rates by 12% in the pilot. The pilot results are Rent the Runway's and Alta's own figures; the feature is now available to all of the rental company's customers.

Eurie Kim, managing partner at Forerunner, framed the bet as a technology threshold: "The fashion industry has been chasing the Clueless closet for 30 years. The problem was never a lack of imagination. The technology simply wasn't good enough." The new money will go to AI infrastructure, growing a team of ten across consumer and enterprise products, and localising the app for other markets. The enterprise line matters: it suggests that even the best-funded consumer apps see selling technology to brands as a more dependable business than consumer subscriptions or commissions alone.

What do design and content tools attract?

The second cluster sells directly to fashion companies. Raspberry AI, founded by Cheryl Liu, a former KKR analyst who later worked at Amazon and DoorDash, raised a USD 24 million Series A led by Andreessen Horowitz in January 2025, about ten months after a USD 4.5 million seed. Its generative AI platform is built for fashion design teams; named customers include Under Armour, Gruppo Teddy and MCM, out of about 70 in total at the time. "There's a lot of terminology behind that sweater that a Midjourney does not know," Liu told TechCrunch, describing why she believes a fashion-specific model can compete with general image generators. The company said it would use the money to hire engineers and sales staff and to expand into home, furniture and cosmetics design, a sign that even fashion-first tool makers look for adjacent markets to justify venture-scale growth.

That argument is being tested. General-purpose image models have improved fast, and the fashion-specific layer has to stay ahead of them. Crunchbase's 2025 list also included planning, sizing, textile inspection and recycling companies, but they raise smaller sums and attract less attention than consumer apps, even though their buyers are the brands that pay most reliably.

Is anyone funding AI for planning, production and materials?

Yes, but in smaller amounts and with less noise. The Crunchbase list of AI fashion startups includes Syrup Tech, which predicts demand for retailers, with USD 25.1 million raised in total; Smartex.ai, a Portuguese company that detects defects in textile production, with USD 27.6 million; and Laws of Motion, which works on precision fit clothing, with USD 10.2 million. Further down the chain sit textile sorting and recycling companies such as Matoha Instrumentation and Refiberd, with seed rounds of a few million at most.

These businesses sell to brands and manufacturers rather than shoppers, their results are measured in stock levels, defect rates and margins, and their investors tend to be more specialised. For fashion companies, they are arguably the more important category, because inventory and production decisions determine most of a company's costs and environmental footprint. Their funding may also prove more durable in a market correction, because their customers pay for measurable savings rather than engagement.

Which fashion-tech companies were sold or shut down?

The clearest verdict on the previous generation of fashion platforms came in April 2025, when Zozo, the Japanese online fashion platform, agreed to buy Lyst for USD 154 million. Lyst, a fashion shopping search platform serving 160 million shoppers and 27,000 brands and retailers, had been valued at USD 700 million in 2021 and had pursued a stock market listing that never happened. Zozo, which reaches 12 million users a year, said it would invest in accelerating Lyst's AI discovery tool, and Lyst could draw on Zozo's sizing and fit technology. Utahiro Inui, Zozo's executive director, said Lyst "has built an exceptional platform that aligns perfectly with our vision". In other words, AI was the rationale for buying a platform at less than a quarter of its former valuation. For the investors who backed Lyst at the peak, the deal crystallised a heavy loss; for Zozo, it bought an international audience of shoppers and brand relationships that a new AI startup would take years and far more money to build. That asymmetry, between the cost of building reach and the cost of buying it after a correction, is something every investor in the new generation of AI shopping apps has to weigh.

Two AI-native startups did not make it to a sale. Alle, an Indian AI stylist that recommended outfits from more than 1,000 brands, raised USD 3 million in seed funding from Elevation Capital and others in December 2023. Its three co-founders, all former executives of the Indian e-commerce company Meesho, decided in October 2025 to close after six pivots failed to find a viable business model; co-founder Prateek Agarwal wrote that "the opportunity cost of everyone's time outweighed another uncertain pivot", while noting that the team had "shipped an AI-first consumer product used by a few millions". Reach, in other words, did not translate into a business. NeuroPixel.AI, an Indian company founded in 2020 and backed by Flipkart Ventures that built virtual try-on, synthetic models and cataloguing tools for fashion e-commerce, announced in April 2026 that it was shutting down its services. Its co-founder Arvind Venugopal Nair said the company "got massively outgunned overnight sometime in late 2025" as general image models such as Google's Nano Banana Pro caught up, and that "the game shifted from IP towards distribution quickly".

What are the risks for investors and for brands?

• Foundation models move the goalposts. NeuroPixel.AI's closure shows that years of proprietary work can be matched by a general model release. Startups whose only edge is image quality are exposed.
• Consumer reach is not revenue. Daydream earns a percentage of each sale it generates, and Phia's coverage says little about its revenue. Download counts and pilot results are company claims, not audited revenue.
• Valuations can reset sharply. Lyst's sale at less than a quarter of its 2021 valuation is a reminder that a platform's AI story does not protect it from the market.
• Data are fragmented. Crunchbase, CB Insights and other trackers use different definitions, so no single reliable total for AI fashion-tech funding exists. Announced figures are sometimes revised, as with differing reports of Phia's round size.
• Dependence cuts both ways. Brands that route customers through an AI stylist gain reach but risk ceding the customer relationship and the data that comes with it.

What should fashion companies do now?

1. Map the startups courting you List the AI stylists, search agents and design tools approaching your teams, with their funding, investors and business models.
2. Ask how each one makes money Commission, subscription or enterprise licence each create different incentives over your products and customer data.
3. Pilot with measurable outcomes Copy the Rent the Runway approach: a limited pilot with metrics such as add-to-bag rate or returns before a full rollout.
4. Test against general models Before paying for a fashion-specific tool, check whether a general model already does the job well enough.
5. Protect your product data Negotiate who owns the catalogue data, imagery and customer insights that flow through partner platforms.
6. Plan for a partner failing Keep exit clauses and data export rights, given how quickly well-funded startups can close or be sold.