Fleek Raises $25 Million to Build AI Infrastructure for Secondhand Fashion

Fleek has raised $25 million in Series B funding to expand the AI infrastructure it has built for the global secondhand clothing industry, an area the company says has kept more than 12 million garments in circulation while saving an estimated 13 billion litres of water and avoiding roughly 23,000 tonnes of carbon dioxide emissions. The round was led by Burda Principal Investments, an early backer of resale platform Vinted, with participation from eBay, FJ Labs and H14 alongside existing investors including Andreessen Horowitz and Y Combinator. The company's platform now connects more than 2,000 verified wholesale suppliers and graders with over 50,000 retailers and resellers across more than 100 countries.

 

Why Secondhand Fashion's Infrastructure Has Lagged Demand

 

The problem Fleek is targeting is a mismatch between growth and infrastructure. Secondhand fashion demand is growing three times faster than the traditional apparel market, yet the company says the supply chain behind a market worth more than $200 billion remains largely manual, fragmented and offline, with garments assessed by hand and traded across disconnected networks with little pricing transparency. Up to 24 billion items move each year from donation bins in cities like London, Paris and New York to textile sorting and grading centres, a volume the company argues cannot scale further without digitisation.

That gap between demand and capacity is the commercial opportunity underlying the funding round. An industry still relying on manual grading struggles to process rising volumes efficiently, which limits how much clothing can be recovered and resold rather than discarded, directly linking the business case to the environmental outcome of keeping more garments in circulation.

 

Read more: Airbus and MTU to Form Joint Venture for Hydrogen Aircraft Engine

 

How the AI Grading System Works

 

At the centre of the platform is Fleek Sort, a vision-language model trained on millions of secondhand marketplace transactions collected over four years. The system identifies, categorises, grades and merchandises garments from photographs or video, converting a process traditionally done by hand into a digital workflow. It is already used by graders in sorting hubs in Pakistan, India and Dubai, with pilots under way in the UK, Europe and the US, and the company says the model continues learning from real transaction outcomes as more inventory passes through it.

Once graded, inventory is listed automatically on Fleek's marketplace, where AI systems handle pricing, search, recommendation and buyer matching. Each transaction generates further data that refines the platform's understanding of secondhand inventory, which the company frames as a proprietary intelligence layer rather than simply a listings platform. That data advantage is central to Fleek's competitive position, since a system trained specifically on secondhand goods, where condition, grading and demand vary enormously, addresses a problem general-purpose commerce platforms are not built to solve.

 

Explore OneStop ESG Marketplace: Sustainable textile

 

The Environmental Case Behind the Business

 

The company frames its environmental impact as a direct function of keeping clothing in use rather than routing it to landfill or incineration. Beyond the 12 million items it says it has helped keep in circulation, Fleek points to higher recovery rates for suppliers and more efficient sourcing for buyers as mechanisms that reduce the volume of wasted clothing overall. Those figures are the company's own estimates rather than independently audited, but they illustrate the scale of impact the business argues is possible once a fragmented, manual supply chain becomes searchable and data-driven.

Chief executive Abhi Arora described the current system for secondhand clothing as broken, pointing to how a garment might travel thousands of miles and be sorted by hand before reaching a vintage shop, if it makes it there at all. Burda's Julian von Eckartsberg compared the opportunity to the firm's early bet on Vinted, framing Fleek's growing supplier network and technology as the infrastructure the next generation of the resale market will depend on. Whether the AI grading system maintains its accuracy as it expands into new markets in the UK, Europe and the US, and whether the platform's growth translates into a proportional increase in garments diverted from waste, will determine whether Fleek's environmental claims scale alongside its commercial ambitions.