OpenAI $1T IPO vs Anthropic vs Google 2004: Historical Take
OpenAI $1T IPO vs Anthropic vs Google 2004: Historical Comparison
OpenAI’s confidential S-1 reportedly targets a $1 trillion+ IPO valuation. Here’s how that compares to history’s biggest tech listings — and what 2026 investors should learn from 2004.
Last verified: June 10, 2026
The historical scoreboard
| Company | IPO date | IPO valuation | IPO revenue (annualized) | P/S at IPO |
|---|---|---|---|---|
| August 2004 | $23B | $3.2B | ~7x | |
| May 2012 | $104B | $4.4B | ~24x | |
| Alibaba | September 2014 | $169B | $8.6B | ~20x |
| Uber | May 2019 | $82B | $11B | ~7x |
| Saudi Aramco | December 2019 | $1.7T | ~$330B | ~5x |
| Snowflake | September 2020 | $33B | $0.4B | ~80x |
| SpaceX | June 12, 2026 | $1.77T | $24B | ~73x |
| Anthropic (private) | May 27, 2026 | $965B | $10–12B | ~85x |
| OpenAI (target IPO) | Q4 2026 / Q1 2027 | $1.0T+ | $24B | ~40x |
Google 2004 — the closest analog
Google’s August 2004 IPO is the historical comparison most cited for OpenAI:
- IPO size: $23B valuation, $1.9B raised
- Revenue at IPO: $3.2B annualized
- YoY growth: ~119% at the time
- Profitable: Yes — $286M net income in H1 2004
- Outcome 5 years later: $200B+ market cap (~9x return)
- Outcome 20 years later: $2T+ market cap (Alphabet)
Why this matters for OpenAI: Google IPO’d profitably with strong unit economics. OpenAI is larger in absolute revenue ($24B vs $3.2B) but smaller in margins and aggressively pricing in future growth. The 2004 outcome was extraordinary — Google grew into and far beyond its valuation. OpenAI is asking investors to pay for an outcome roughly 40x richer than Google’s IPO multiple.
Facebook 2012 — the cautionary tale
Facebook IPO’d at $104B and a 24x P/S multiple — a meaningful step up from Google’s. The first year was rough:
- Day 1: $38/share open, closed at $38.23 (flat)
- Three months later: $17.55 (-54% from IPO)
- One year later: $24.63 (-35% from IPO)
- 18 months later: recovered to IPO price
- Today (2026): $700+ (~18x return on IPO buyers willing to hold)
Why this matters: rich IPOs can spend a long time underwater before fundamentals catch up. OpenAI’s 40x P/S target leaves room for similar pain if growth slows or competitive dynamics shift.
Saudi Aramco 2019 — the absolute-size comp
The current record-holder for largest IPO ever (until SpaceX in June 2026). Aramco listed at $1.7T valuation — strikingly close to SpaceX’s June 2026 $1.77T pricing.
Why this matters: trillion-dollar IPOs are technically possible. They typically require:
- Massive existing revenue base
- Anchor sovereign or strategic investors
- Limited float (small percentage of shares offered)
- Strong narrative supporting the premium
OpenAI and Anthropic don’t have Aramco’s revenue ($330B). They’re banking on narrative and growth.
The 2026 trio side-by-side
| Property | SpaceX (SPCX) | Anthropic | OpenAI |
|---|---|---|---|
| IPO valuation | $1.77T | $1.0–1.2T target | $1.0T+ target |
| Revenue | $24B | $10–12B | $24B |
| P/S | 73x | 85x | 40x |
| Profitability | Cash-flow neutral | Cash-flow negative | Cash-flow negative |
| Growth YoY | 80% | 200%+ | 150% |
| Path to IPO | Aug 2025 → June 2026 | June 2026 filing | June 2026 filing |
| Market debut | June 12, 2026 | ~October 2026 | ~Q4 2026 / Q1 2027 |
| Capital raised at IPO | $75B | ~$50–80B (estimated) | ~$60–100B (estimated) |
| Comparable historical IPO | Aramco 2019 | Snowflake 2020 | Facebook 2012 |
What’s actually different about 2026
1. Multiple trillion-dollar IPOs in 6 months
Aramco was a one-off in 2019. 2026 has SpaceX in June, Anthropic in October, OpenAI in Q4/Q1. This concentrates IPO supply at an unprecedented scale.
2. Pre-IPO valuation already exceeded most public companies
Anthropic at $965B private valuation would be the 7th most valuable public company in the world if listed today (between Berkshire and Tesla). OpenAI at $852B private would be ~10th.
3. AI as a category didn’t exist in 2004
Google IPO’d into a known category (search advertising). OpenAI is IPO’ing into a category it largely invented. Categorical TAM estimates range from $1T to $10T+ by 2030.
4. Hyperscaler dependence
None of the 2026 trio is fully vertically integrated. SpaceX is closest (Falcon + Starlink + Colossus). OpenAI depends on Microsoft Azure. Anthropic depends on AWS + Google Cloud. Investors need to model not just the AI lab but the hyperscaler relationship.
The 1999 comparison — apt or alarmist?
Similarities to 1999:
- Multiple trillion-dollar tech IPOs in months
- Retail mania (commission-free apps, fractional shares)
- Narrative-driven valuations ahead of fundamentals
- “This time it’s different” arguments
Differences from 1999:
- 2026 leaders have real revenue ($10B–$24B) — Pets.com had $0
- Hyperscalers (AWS, Azure, GCP) own the infrastructure layer
- AI usage is broadly real (ChatGPT 800M+ WAU, Codex 5M+ developers)
- Hardware constraint (GPUs, power) creates natural scarcity
Net: the 2026 AI wave has real fundamentals but is priced for perfection. A 30–50% drawdown post-IPO would not be historically unusual even if the long-term trajectory is positive.
What to actually do
If you’re bullish on AI generally
- 5–10% portfolio basket across SpaceX, Anthropic, OpenAI plus Nvidia, AMD, Broadcom, hyperscalers
- Equal-weight the AI labs if you can’t pick a winner
- Buy 30–90 days post-IPO to avoid speculative froth (and pre-lockup-expiry pressure)
- Limit single-name exposure to 2–3% of portfolio per AI lab
If you’re skeptical
- Wait 12–24 months for fundamental confirmation
- Watch for: GPT-6 delivery, Anthropic enterprise growth, EU AI Act impact, open-source frontier
- Re-evaluate at 50% drawdown if it occurs
- A 30% drawdown without fundamental degradation could be a great long-term entry
If you’re a day trader
- Statistical evidence: retail loses money on day-one IPO trading
- IPO premiums (40–80% over IPO price on day one) are usually unsustainable
- Wait for the 30-day rolling average to stabilize
What could go wrong
| Scenario | Effect on 2026 trio |
|---|---|
| GPT-6 disappoints or delays | OpenAI -30%+, Anthropic +10% relative |
| Llama 5 / Qwen 4 closes frontier gap | All AI labs -20–40% |
| EU AI Act August 2 strict enforcement | OpenAI/Anthropic European revenue at risk |
| Compute scarcity or power-grid limits | All three constrained |
| Macro recession / rate spike | High-duration assets -30–50% |
| Major safety incident (model-driven harm) | Entire category re-priced |
| Antitrust / DOJ action | OpenAI most exposed |
| Starship launch failure | SpaceX -20–30% |
| Musk distraction or scandal | SpaceX -15–25% |
Related reading
- OpenAI S-1 filing vs Anthropic IPO race
- OpenAI $852B vs Anthropic $965B vs SpaceX $1.77T
- SpaceX IPO $1.77T vs Saudi Aramco 2019
- Anthropic IPO vs Google 2004 vs Facebook 2012
- SpaceX IPO June 11 2026: How to invest SPCX
- Anthropic IPO October 2026 timeline
Sources
- The Guardian: OpenAI IPO files for public stock market (June 8, 2026)
- CBS News: OpenAI files confidential initial public offering
- Business Insider: OpenAI files confidential S-1
- TechFundingNews: OpenAI reportedly eyes $1T valuation
- NBC News: Anthropic files for IPO before OpenAI
- Yahoo Finance: $960 Billion Anthropic Beats OpenAI to the IPO Filing
- CNBC: SpaceX targets $135 IPO price at valuation of $1.77 trillion
- NYT: SpaceX IPO to Be Largest Ever
- Historical SEC filings: Google (2004), Facebook (2012), Alibaba (2014)