The AI IPO pipeline is the most anticipated in tech history. Databricks valued at $62 billion. Anthropic at $61.5 billion. Stripe processing $1 trillion a year with AI-powered fraud detection. Together they represent over $200 billion in private value waiting to hit public markets. ## The AI IPO Pipeline: Who's Likely Going Public
Tier 1: Near-Certain 2026-2027 Candidates
| Company | Last Valuation | Est. Revenue (2025) | Likely Route | Why Now |
|---|---|---|---|---|
| Databricks | $62B | $2.4B+ ARR | Traditional IPO | Revenue scale + path to profit |
| Stripe | $91B (internal) | $4B+ revenue | Direct listing | Doesn't need cash, wants liquidity |
| Anthropic | $61.5B | $900M+ ARR | Traditional IPO | Needs massive capital for compute |
| CoreWeave | $35B | $1.9B+ revenue | IPO (filed S-1) | Heavy debt, needs equity balance |
| Canva | $26B | $2.5B+ ARR | Direct listing or IPO | Profitable, employees want liquidity |
| Cerebras | $4B+ | $500M+ revenue | Traditional IPO | Filed S-1 in late 2025 |
CoreWeave filed its S-1 in early 2026 and will be the first major AI infrastructure pure-play to test public markets. How it prices will set the benchmark for every AI infrastructure company behind it.
Tier 2: Probable 2026-2027 Candidates
| Company | Last Valuation | Key Metric | Route Consideration |
| Scale AI | $14.3B | $1B+ ARR | IPO after government contract growth |
| Anduril | $14B+ | $1B+ backlog | Traditional IPO, strong defense appetite |
| Figma | $12.5B | $700M+ ARR | IPO after Adobe deal fell apart |
| Plaid | $6.5B | $500M+ revenue | IPO, fintech-AI convergence |
| Wiz | Acquired ($32B) | N/A | Could do a subsidiary IPO under Alphabet |
Tier 3: Watching the Market
OpenAI is the elephant in the room. At a $300 billion private valuation and roughly $4 billion in annualized revenue, a traditional IPO would be the biggest tech offering since Alibaba. But their ongoing restructuring from nonprofit to for-profit creates legal complexity. Most analysts expect 2027-2028 for an OpenAI public offering, probably through a holding company structure.
Routes to Public Markets
Traditional IPO
How it works: File S-1, do a roadshow, banks price and sell shares to institutions. Company raises fresh capital.
Best for: AI companies that need money. Anthropic needs billions for compute. CoreWeave needs equity to offset $7.5B+ in debt. Traditional IPOs raise $500M-$5B while creating a public currency for acquisitions and comp.
Cost: 5-7% underwriting fees plus $2-5M legal/accounting. A $2B IPO costs roughly $110-145M in fees.
Recent AI example: Reddit's March 2024 IPO raised $748 million at $6.4 billion. Despite shaky early trading, by early 2026 Reddit's cap topped $30 billion — 4x+ returns for IPO investors.
Direct Listing
How it works: Existing shares go on a public exchange. No new capital raised. The market discovers the price on day one.
Best for: Profitable AI companies that don't need cash but want to give employees and early investors liquidity. Stripe and Canva are textbook candidates — both reportedly profitable with no urgent capital needs.
Cost: Much cheaper. No underwriting fees (saves 5-7%). Advisory runs $10-30M versus $100M+ for a traditional IPO.
Track record: Spotify (2018), Slack (2019), Coinbase (2021), Roblox (2021) all went this route. Mixed results — Spotify and Coinbase worked well; Slack ended up selling to Salesforce below its listing peak.
SPAC Mergers
How it works: A blank-check company raises capital through its own IPO, then merges with a private company to take it public.
2025-2026 reality: SPACs cooled dramatically from the 2021 frenzy. Back then, 613 SPACs raised $162 billion. In 2025, fewer than 50 raised under $10 billion. But AI-focused SPACs are seeing a second wind.
Best for: AI companies in the $1-5 billion range that want to go public faster. SPACs close in 3-6 months and let the target share forward projections (not allowed in traditional IPOs).
| SPAC Characteristic | 2021 Peak | Current 2026 |
| Average size | $300M | $150-200M |
| Time to merge | 3-6 months | 4-8 months |
| Redemption rate | 20-40% | 60-80% (investors more cautious) |
| Founder promote | 20% | 10-15% (reduced) |
| Target valuation | $1-20B | $500M-$5B |
A handful of AI-focused SPACs formed in late 2025 and early 2026, targeting defense AI, healthcare AI, and AI infrastructure — verticals where the SPAC sponsors bring real domain expertise.
Financial Benchmarks for AI IPOs
Public markets are pickier than private markets. For AI companies, the bar is higher — investors want Rule of 60+ for premium pricing.
| Scenario | Growth | Margin | Score | Market Response |
| Hypergrowth AI | 100% | -40% | 60 | Premium |
| Growth AI | 60% | -10% | 50 | Fair |
| Mature AI | 30% | +15% | 45 | Standard |
| Struggling AI | 20% | -20% | 0 | Discounted or delayed |
Databricks reportedly hits 60%+ growth with improving margins — firmly in premium territory. CoreWeave's revenue grows 100%+ but infrastructure costs compress margins, making its Rule of 40 score dependent on how the market reads gross vs. operating margins.
How to Buy Pre-IPO AI Shares
Secondary Market Platforms
For accredited investors, secondary markets offer a way in before the bell rings:
Forge Global (FRGE) — Largest secondary marketplace. $15B+ in transactions across 500+ companies. Anthropic, Databricks, and Stripe shares trade regularly. Minimum usually $100,000.
EquityZen — Lower minimums ($10K-$50K). Buys shares from employees and early investors through SPVs. Has helped OpenAI, Canva, and Stripe pre-IPO trades.
Hiive — Newer platform with a bid-ask model like a stock exchange. Real-time pricing, lower fees (3-5% vs. Forge's 5-7%). Strong demand for AI shares.
CartaX (Carta) — Connects shareholders directly with buyers, using Carta's cap table data on 40,000+ companies.
Secondary Market Pricing
Pre-IPO AI shares typically trade at 10-30% below the last primary round. That discount reflects illiquidity, information gaps, and transfer restrictions.
But for hot AI companies, shares sometimes trade at a premium. OpenAI shares reportedly traded 15-25% above the $300 billion mark in early 2026.
| Company | Last Round | Secondary Premium/Discount | Typical Lot |
| OpenAI | $300B | +15-25% premium | $250K+ |
| Anthropic | $61.5B | +5-10% premium | $100K+ |
| Databricks | $62B | Flat to -5% | $100K+ |
| Stripe | $91B (internal) | -10-15% discount | $100K+ |
| CoreWeave | $35B | -5-15% discount | $50K+ |
Pre-IPO Risks
Lockups. After IPO, secondary buyers typically face 90-180 day lockups. Stocks can drop hard when lockup expires and selling pressure spikes.
Valuation compression. Private valuations don't always hold in public markets. Even strong companies sometimes see 20-40% compression from last private round to IPO price.
Stuck capital. If the company delays or scraps the IPO, your shares get even more illiquid. Some investors in late-stage companies have waited 5-7 years for liquidity.
2026 Predictions
CoreWeave sets the tone. First major AI infrastructure company to test public markets. A strong debut opens floodgates. A weak one delays everything.
Databricks likely files by mid-2026. $2.4B+ ARR with strong growth — the most obvious IPO candidate. CEO Ali Ghodsi has publicly discussed readiness. Expect a $60-80 billion target.
3-5 AI companies reach public markets in 2026. Beyond CoreWeave and likely Databricks, expect a few more via traditional IPOs, direct listings, or SPACs.
The AI premium holds for leaders but fades for followers. First movers get novelty and scarcity premiums. Later entrants face tougher scrutiny as public investors build better AI valuation frameworks.
For investors, the 2026 AI IPO wave is a generational opportunity. The key is separating companies with durable AI advantages from those riding temporary hype — and getting positioned before the window opens.
