Bending Spoons Seeks $1.62B IPO With AI Writing 90% of Its Code

The most striking item in Bending Spoons' public filing is not the revenue trajectory. It is that the Milan-based acquirer of internet brands put a specific number on AI-driven code automation inside an SEC document: more than 90% of the company's pull requests are now authored or co-authored by AI, up from less than 10% in Q1 2025, with around 70% authored by AI alone. According to Reuters, the company is seeking to raise $1.62 billion in a US IPO, offering 58 million shares at $26 to $28 each, which at the top of the range implies a valuation of around $19 billion.

Bending Spoons was founded in 2013 and operates as a private-equity-style roll-up of consumer internet businesses, including Vimeo, AOL, Eventbrite, Evernote, and WeTransfer. The company has described itself as buying ailing internet brands with the hope of turning them profitable. The AI story is where the model has become distinctive: revenue per full-time employee reportedly rose from $1.12 million in 2023 to $2.57 million in 2025. For the first three months of 2026, the company posted net income of $27.5 million on $601 million in revenue, compared with a net loss of $112 million on $259 million in revenue for the same period a year earlier.

What makes this IPO worth watching beyond the usual roll-up story is that the S-1 is a signed SEC disclosure, not a press release. The AI productivity claims are filed and will be revisited every quarter. That creates a live public data series on what aggressive AI adoption actually does to operating leverage over time, something the industry has had plenty of assertions about but very few auditable examples of at this scale.

What the reporting does not give you is the underlying headcount, a breakdown of which AI tools generate those pull requests, or any disclosure of how code quality and security are monitored at this automation level. The valuation also has its own caveat: Bending Spoons raised at an $11 billion valuation as recently as last year, and the public market price at the top of the range represents a meaningful step up from that.

If the offering prices well and the stock holds in the aftermarket, it will create a publicly-traded benchmark for AI-native operations inside an acquisition-driven business, one that any operator or investor considering a similar playbook will be watching on a quarterly basis.