Analysis
Two days after Crunchbase identified eight startups as the clearest candidates for a narrowing 2026 IPO window -- Anthropic, Oura, Kraken, SambaNova, Notion, Stripe, Stegra and OpenEvidence -- exactly one of them has actually converted from candidate to filed public offering: Oura, which submitted its S-1 on Sept. 3.
What Changed, and What Didn't
Oura's filing is a genuine status change, not just incremental progress -- it moved from "eyeing a September IPO" to an actual registered public document with audited revenue, membership and risk-factor disclosures. Pulse covered the filing's specifics separately, including 74% revenue growth to $1.21 billion and 5 million paid members. None of the other seven names on Crunchbase's list have made an equivalent move in the same window.
“Several names on this list are still working out exactly what story their own S-1 will need to tell.”
Anthropic remains the most advanced of the rest, having confidentially filed with the SEC back on June 1 -- a real step, but one that predates this specific list by two months and hasn't progressed to a public S-1 in the time since. Kraken's co-CEO described the company as "about 80% ready" for a 2026 listing as of the most recent public comment Pulse has on record, but readiness commentary is not the same as an actual filing, and crypto-exchange IPOs specifically have a documented history of slipping timelines when market conditions turn less favorable. SambaNova's CEO has said the company is "strongly considering" a US listing at a valuation anchored to its $11 billion Series F post-money mark, softer language than anything approaching a committed date. Notion's recent board appointments and enterprise AI revenue growth are being read as readiness signals without an actual timeline attached. Stripe's own CEO, Patrick Collison, has said the company is "in no rush" -- the clearest signal on the entire list that inclusion in a "candidates to watch" roundup doesn't mean a 2026 listing is actually likely. Stegra, the green-steel manufacturer, and OpenEvidence, the doctor-facing AI platform whose own IPO timing reportedly follows behind the foundation-model companies it depends on, round out the list with the least near-term visibility of the eight.
Why the Gap Is the Story
The pattern here rhymes with what Pulse observed separately this week comparing AI-infrastructure IPO pace to AI-lab IPO pace: companies with legible, analyzable revenue -- Oura's subscription-hardware model, in this case -- are moving faster through the actual filing process than companies whose valuation depends more heavily on growth narrative, competitive dynamics that can shift overnight, or safety and regulatory questions still being actively litigated in public. Oura had real, auditable growth numbers to put in a prospectus. Several names on this list are still working out exactly what story their own S-1 will need to tell.
What to Watch
The most useful update to this list won't be another company merely reaffirming interest -- it'll be the next name to file an actual public S-1, converting from candidate to registrant the same way Oura just did. Until that happens, the honest read on this list is that it names eight plausible companies with wildly different actual timelines, and treating all eight as equally likely to list in 2026 already looks like it was too generous two days after the list was published.
