Six companies dominate the conversation about what goes public next and each one tells a different statistical story about growth, risk and timing.
- SpaceX — Public (NYSE: SPCX), listed June 2026. Valuation: ~$2.4B (2012) → ~$2.1T today.
- Anduril — Private, no IPO date set. Valuation: ~$1B (2019) → $61B (May 2026).
- OpenAI — Private, S-1 filed, targeting 2027. Valuation: ~$20B (2021) → $852B (March 2026).
- Anthropic — Private, S-1 filed, targeting October 2026. Valuation: ~$5B (2022) → $965B (May 2026).
- Databricks — Private, targeting 2027 or later. Valuation: ~$6.2B (2019) → $190B (August 2026).
- Deep Fission — Public (Nasdaq: FISN), filing complete. Valuation: founded 2023 → ~$1.66B IPO target.
SpaceX is the only company on this list already trading. Its stock sits near $155 a share, a ~$2.1T market cap up from a $135 IPO price and a 52-week range of $105 to $226 that shows how volatile the first three months of trading have been. Full-year 2025 revenue was $18.7B, with Starlink contributing $11.4B of that, and Q2 2026 revenue reached $7.8B, up 92% year over year. Starlink subscribers passed 12M by Q2 2026 even as average revenue per subscriber fell from $99 in 2023 to $66 in early 2026, the cost of expanding into price-sensitive markets in Africa, Southeast Asia and Latin America. SpaceX employs 22,621 people, launched 165 Falcon 9 missions in 2025 (about 51% of all global orbital launches) and has flown 12 integrated Starship tests, four of them full successes. Competition: Amazon’s Leo formerly Project Kuiper is the most-watched rival — it entered enterprise beta in April 2026, but as of August 2026 still had zero active commercial subscribers, with full deployment not expected until 2027–2029. Also notable: Starlink struck an $11.57B satellite deal that reportedly handed its “only real rival” an unexpected shortcut, worth watching since it could accelerate Kuiper’s timeline. Bottom line: the $2T price is largely a bet on Starship and future AI/data-center revenue rather than current fundamentals. Starlink is still growing subscribers fast but at declining per-user economics and the stock has already shown it can swing ±$100B in a day on sentiment alone.
Anduril has grown from a $1B unicorn valuation in 2019 to $61B after its May 2026 Series H, though a newer funding round reportedly targets just $28B, a discrepancy worth watching. The company is projecting $4.3B in 2026 revenue against a projected $1.2B operating loss on roughly 7,000 employees, spending heavily ahead of revenue recognition on a March 2026 US Army contract worth up to $20B over ten years. Its CEO has publicly warned against IPOing “in the middle of a hype cycle,” leaving the timing genuinely open.
OpenAI has raised $176.4B in total capital and reached an $852B valuation by February 2026, up from roughly $20B in 2021, with bankers discussing up to $1T for an eventual listing. Annualized revenue topped $25B by February 2026 and monthly revenue hit $2B in March, on a lean 4,500 employees relative to its user base: ChatGPT crossed 900M weekly active users in February 2026 and 1B monthly active users by June, the fastest app in history to reach that scale, with 1M+ business customers and 50M+ consumer subscribers. Its CFO has said the company will be public “in 2027,” possibly sooner.
Anthropic shows the fastest revenue trajectory of the group, climbing from a $1B annualized run rate in December 2024 to $47B in May 2026 and $65B by July 2026. Its valuation rose from roughly $5B in a 2022 Google-backed round to $965B by May 2026, with bankers discussing a listing valuation up to $2T against an October 2026 target date, the closest of any private company here to an actual listing. More than 300,000 businesses use Claude, over 1,000 of them spending above $1M a year and roughly 70% of the Fortune 100 are customers.
Databricks has grown from a $6.2B valuation in 2019 to $190B after an August 2026 strategic round, with revenue climbing from $5.4B to $7B in annualized run rate across the first two quarters of 2026, growth above 80% year over year. It serves roughly 15,000 customers, 650 of them spending $1M or more annually, with net dollar retention above 140% on a headcount estimated between 8,000 and 10,000. Its CEO has explicitly ruled out a 2026 listing, calling it “a terrible year to go public,” pointing instead to 2027 or later.
Deep Fission is the outlier of the group: a small, already-public, pre-revenue nuclear startup rather than a scaled private giant weighing its IPO timing. It employs just 62 people, held $93.0M in cash as of June 2026, posted a $61.0M net loss in 2025 and carries a going-concern warning attached to a $31M loss in its most recent quarter. Its IPO targets a $1.66B valuation on a $156M raise and the company estimates it needs at least $200M more before its first commercial reactor, targeted for Kansas in 2027, can go live, despite 18.5 GW in customer letters of intent.
Taken together the six split into three groups by statistical profile: one already-public and volatile (SpaceX), four large and fast-growing but still cash-burning private companies weighing a 2026 or 2027 listing (Anthropic closest, Databricks furthest out, OpenAI and Anduril in between), and one small, already-public, pre-revenue company carrying real near-term cash risk (Deep Fission).