Oura's rumored September IPO is being priced on AI, not the ring

The one thing Oura has stated on the record is that it made a confidential S-1 submission on May 21. The press release at the time said the number of shares to be offered and the price range hadn't been set yet, and that the company would go public once the SEC review wrapped up and market conditions allowed. A public registration statement laying out actual offering terms still hasn't appeared.

Timing	Round / event	Valuation
December 2024	Series D — $200 million raised	$5.2 billion
October 2025	$900 million raised	~$11 billion
May 2026	Confidential S-1 filed	Undisclosed
September 2026 (rumored)	Up to $3 billion offering	Above $16 billion

The pace of the valuation climb is steep. Last October's round, led by Fidelity with new investor ICONIQ joining in, valued the company at roughly $11 billion. Hitting $16 billion would mean the valuation grew 1.5x in just eleven months. Revenue has climbed just as fast. 2024 revenue topped $500 million, double the year before, and the company has said it expects another doubling to $1 billion in 2025. CEO Tom Hale told CNBC last November that he expects 2026 revenue to come in near $2 billion. That said, the company still hasn't published finalized 2025 results.

The valuation is built on software, not the ring itself

If you look at Oura purely as a hardware company, these numbers don't quite add up. What the company has actually been building over the past two years sits less in the sensors inside the ring and more in the AI layered on top of them.

The flagship product is Oura Advisor, a chat-based health coach that launched in beta in July 2024 and rolled out to all members in March 2025. It reads a user's sleep, activity, and stress data and answers questions directly. The company hasn't said whose model powers it, describing it only as a state-of-the-art large language model.

In February, Oura released its first proprietary model — a large language model built specifically for women's health that runs inside Advisor. Oura said the model runs only on infrastructure it controls, and that member conversations aren't sent out to third parties or used to train public AI systems. Its acquisitions have followed the same trajectory. In 2024 it bought the metabolic health app Veri and Sparta Science, a platform for enterprise physical-data analytics. This March it acquired the gesture-recognition startup Doublepoint, and in April it bought Galen AI, which combines medical records, lab results, medication history, and wearable data into one system. None of the four deal amounts were disclosed.

The most recent signal came just five days ago. On August 20, Oura named Sanjay Chandra as its first chief information officer and, the same day, brought on Han Chiu as senior vice president of AI and software engineering. Ahead of a potential IPO, the company has effectively built a separate command structure for its AI organization.

The lawsuit, too, is really about an AI estimate

The same week brought news in the opposite direction. On the 20th, a class-action lawsuit was filed in the US District Court for the Northern District of California. The plaintiffs argue that Oura advertised sleep-stage accuracy rates of 79%, and more recently 95%, even though the ring can't capture the scalp-electrode or eye-movement signals needed to determine sleep stages in the first place. The filing goes so far as to describe the AI's estimates as no better than a coin flip.

Oura has pushed back, saying that like other sleep wearables, it estimates sleep stages using a combination of physiological signals — heart rate, heart rate variability, movement, breathing patterns, and body temperature — and that multiple independent third-party studies back up its accuracy claims. That said, the 2024 Brigham and Women's Hospital study the company frequently cites was based on just 35 participants over a single night, was funded by Oura, and its corresponding author sits on Oura's medical advisory board. In that study, Cohen's kappa for four-stage sleep classification came out to 0.65 for Oura, 0.60 for Apple Watch, and 0.55 for Fitbit. The 2025 study cited in the lawsuit put four-stage accuracy at just 53.18%.

Korea got its launch the very same day

As it happens, official sales of the Oura Ring 5 began in Korea today, following the timeline announced at a press event at Lotte World Tower in Seoul on the 18th. Pricing runs 630,000 won or 780,000 won depending on color, with the charging case sold separately for 160,000 won. Subscriptions are free for the first month, then 9,400 won a month or 110,000 won a year. Oura has opened 34 offline service centers and a Korean-language call center, but it's operating through a local partner rather than setting up a Korean subsidiary.

The company has made clear it won't adjust its subscription model for the Korean market. George Abbott, who leads Oura's Europe, Middle East, Africa, and Asia-Pacific business, told the Korea Herald: "The short answer is no. We believe in that model, and we're not going to change it for the Korean market." The comparison point is obvious. Samsung's Galaxy Ring retails in Korea for 499,400 won, with no subscription fee and no separate charging case to buy. Even against Oura's cheapest option, the hardware price gap runs about 130,000 won, and once you add the charging case and a first year of subscription, the total gap widens to around 300,000 won.

The regulatory backdrop is also new. Under South Korea's Digital Medical Products Act, which took effect in January last year, a registration and certification system for digital medical and health-support devices began operating this past January. The first product registered under it was Samsung Health, covering functions built into the Galaxy Watch8 and Galaxy Ring, while the first standalone smart ring to register was VIVA Health's VIVA Ring in February. Oura's Korea launch announcement made no mention of this registration system.

Editor's take

$16 billion isn't a number you can explain by counting rings sold. Cumulative sales were last reported at 5.5 million units back in September, and that same figure came up again at this week's Korea press event. What the company keeps pointing to instead is more than 5 million paying subscribers across 150 countries, with renewal rates above 80% after the first year. What holds that renewal rate together is Advisor and the proprietary models behind it. Oura isn't really selling the numbers its sensors capture — it's selling the interpretation of those numbers, and every interpretation is the output of an estimation model.

That's what makes the timing of this lawsuit so uncomfortable. It's not challenging a defective ring; it's challenging the reliability of what the AI estimates. If the accuracy claims wobble, it's not just one feature that's at stake — it's the entire logic that's been used to justify the subscription. Whoop's blood-pressure feature drew an FDA warning letter last July that wasn't resolved until this past June, and that's a case from the same family of problems.

There are two things worth watching in Korea. One is whether subscription-based AI health services can actually work with Korean consumers, in a market where the Galaxy Ring is already cheaper and comes with no subscription — yet Oura has arrived charging both a hardware price and a monthly fee. The other is regulatory. Seven months after Korea opened its registration category for digital medical and health-support devices, the only known registrants — Samsung Health and VIVA Ring — are both domestic. Whether foreign devices start using this category or not will end up shaping how AI health products are allowed to market themselves in Korea going forward.

Whether the IPO happens in September or gets pushed back, the story Oura is bringing to the market isn't about the ring — it's about AI. And the place where that story gets priced isn't just the Nasdaq. It's also a courtroom in the Northern District of California.