Revolut hits $115bn valuation and cuts AI coding costs 90%
Two Revolut stories landed on the same day this week, and both are worth tracking if you follow fintech or enterprise AI cost management.
The $115bn valuation
Revolut completed a secondary share sale that puts its valuation at $115bn, a 53% jump from the $75bn set by a comparable sale a year earlier. Glade Brook Capital Partners, Coatue, Durable Capital Partners, T. Rowe Price, and Baillie Gifford led the round. Existing and new investors joined them.
No new capital entered the company. Employees and early backers sold existing shares at roughly $2,017 each, according to Tech.eu. Revolut did not disclose how much stock changed hands in total.
The $115bn figure is the same price first reported in July and the valuation underpinning CEO Nik Storonsky’s share award announced in August.
“Every bank we launch means more people get a better way to manage their money, wherever they live. Our team makes that happen, and employee ownership has been central to our culture from day one.” — Nik Storonsky, CEO and founder, Revolut
At that valuation, private investors now price Revolut above Barclays and Société Générale, according to Reuters. The contrast with traditional banks is sharp in one metric: Revolut held £2.2bn in loans at end of 2025, a loan-to-deposit ratio of 6%, compared to 55% at HSBC.
Reuters also noted two regulatory setbacks. Revolut was fined in Lithuania for failing to prevent money laundering. In September, it sent customer data to scammers posing as officials. Revolut said no confirmed money laundering was found and that its systems and customer funds were not affected.
The company reported a 38% profit margin on $6bn of revenue in 2025. A stock market listing would come no earlier than 2028. Storonsky has weighed a dual listing on the London Stock Exchange and Nasdaq. Revolut now has more than 80 million retail customers and has launched banks in the UK, Mexico, Australia, and New Zealand this year. It holds banking licences in France, Colombia, and Argentina, and has conditional approval from the US Office of the Comptroller of the Currency for a national bank it plans to open in the first half of 2027.
The AI coding cost story
The valuation news closed a day after The Register reported how Revolut cut its AI coding bill by a factor of ten. The details came from a talk by David Abitbol, who leads AI operations at Revolut, at the AI Engineer conference in New York on 13 October.
“We were highly skeptical about the cost of AI.” — David Abitbol, Revolut, at AI Engineer
Revolut now runs GLM-5.3, a Chinese open-weight model from Z.ai, on its own GPUs. It serves as the default coding model for roughly 2,000 engineers, running on 56 Nvidia B200 chips. Abitbol put the blended cost at 30 cents per million tokens. That puts Revolut’s current invoice at a tenth of what it previously spent with Anthropic and OpenAI.
The path there: Revolut rolled out Claude Code widely in January with strict per-engineer budget limits. Some engineers had been spending between $1,000 and $10,000 a month without a clear return. Self-hosting started with GLM-5.1 in April, and the company has since moved to GLM-5.3.
Revolut’s internal data shows coding time improved by 60% and lead time, from picking up a feature to having it ready to deploy, improved by 70%. The conference’s own write-up notes that the chart does not separate the model change from other contributing factors.
Revolut has not dropped US lab models entirely. Abitbol called the latest Codex and Claude Code models “fantastic” but said they are reserved for exceptions. Image tasks go to Gemini or GPT Luna, since GLM handles text only.
“Open source is now a priority at Revolut, and our economics are favorable. The billing is flat, we don’t have any bill shock at the end of the month, and the frontier is the exception.” — David Abitbol, Revolut
On running a Chinese model inside an EU-regulated bank, Abitbol said Revolut downloads the weights and runs them on infrastructure it controls. No requests go to an endpoint in China, he said.
I’m founder of BizStack at Cagri Sarigoz LLC and a passionate advocate for entrepreneurs.
With over 14 years in tech, marketing, and AI, including my role as Head of SEO at CitizenShipper and co-founder of TaleBot at Intale AI, I’m dedicated to sharing genuine, useful product insights and tips.
At BizStack, I aim to cut through the digital noise to provide clear, actionable advice.
And more than all else, I’m a father to a (always) little girl and a husband.
Contact me at [email protected] for assistance.
