Anthropic has eclipsed OpenAI in global LLM revenue share, is approaching a ~$900B–$1T implied valuation on secondary markets, and is actively preparing for what could be the largest technology IPO in history — potentially as early as October 2026. The bull case rests on enterprise pricing power, infrastructure moat, and a faster path to profitability than any peer.
From $61.5B in March 2025, to $183B at the Series F in September 2025, to $380B at the Series G in February 2026, to an implied ~$900B on secondary markets by May 2026 — a 50× valuation increase in 28 months, with no comparable historical precedent in private markets.
| Round | Date | Amount Raised | Post-Money Valuation | Lead Investors |
|---|---|---|---|---|
| Series E | March 2025 | — | $61.5B | Lightspeed |
| Series F | September 2025 | $13.0B | $183.0B | Lightspeed, Salesforce, Cisco |
| Series G | February 2026 | $30.0B | $380.0B | GIC + Coatue (lead), D.E. Shaw, Dragoneer, Founders Fund, ICONIQ, MGX |
| Series H In Talks | May 2026 | $40–50B | ~$900B | Preemptive offers — board decision pending |
The Series G was led by GIC and Coatue, co-led by D.E. Shaw Ventures, Dragoneer, Founders Fund, ICONIQ, and MGX, with participation from BlackRock, Fidelity, General Catalyst, Sequoia, Temasek, and TPG — at a $380B post-money valuation. Subsequent preemptive offers from investors have been filed in the $850B–$900B valuation range, with the company receiving multiple unsolicited bids totaling $50B.
Three of the world's largest sovereign wealth funds have already anchored Anthropic's investor base — and the broader SWF universe is the most-watched marginal buyer for any Series H or pre-IPO secondary tender, given that frontier-AI cheques sized at $5B+ are increasingly only writable by state-backed pools.
| Sovereign Wealth Fund | Country | AUM | Anthropic Status | Frontier-AI Track Record |
|---|---|---|---|---|
| GIC | 🇸🇬 Singapore | ~$770B | Series G lead | Anchor LP across major late-stage tech rounds; AI infrastructure capex partner |
| MGX | 🇦🇪 UAE — Abu Dhabi | ~$100B | Series G co-lead | Sovereign AI investment vehicle (Mubadala / IHC / AI71); xAI, OpenAI exposure |
| Temasek | 🇸🇬 Singapore | ~$380B | Series G participant | Long-standing AI / enterprise SaaS investor; Stripe, Databricks, Mistral exposure |
| Public Investment Fund (PIF) | 🇸🇦 Saudi Arabia | ~$925B | Not disclosed | HUMAIN sovereign-AI vehicle; xAI Memphis compute deals; major frontier-model capex |
| Mubadala | 🇦🇪 UAE — Abu Dhabi | ~$330B | Not disclosed | MGX co-shareholder; data-center + AI infra co-investments with G42, Microsoft |
| ADIA | 🇦🇪 UAE — Abu Dhabi | ~$1.05T | Not disclosed | Generally tech-active via fund-of-funds; discrete direct private exposure |
| QIA | 🇶🇦 Qatar | ~$525B | Not disclosed | Late-stage tech via Coatue, Tiger Global, Insight Partners fund commitments |
| NBIM (Norway) | 🇳🇴 Norway | ~$1.85T | Public-equities only | Largest SWF globally; mandate limits direct private placements — AI exposure via AMZN / GOOGL / MSFT public stakes |
| CIC | 🇨🇳 China | ~$1.35T | Geopolitical headwind | CFIUS scrutiny + US export-control regime structurally limits frontier-AI direct exposure |
| KIA | 🇰🇼 Kuwait | ~$1.0T | Not disclosed | Discrete late-stage tech allocations; smaller direct AI footprint than UAE / Singapore peers |
Read-through: the marginal buyer of a Series H or IPO cross-over book is a Gulf / Singapore SWF. AUMs above are public 2024–2025 disclosures; AI / private-deal allocations are a small fraction of total assets but represent the deepest pool of remaining late-stage capital at frontier-model sizing. The Anthropic Series H board decision will likely involve direct conversations with PIF, ADIA, and Mubadala alongside the three SWFs already on the cap table.
Annualized revenue surged from $87M in January 2024 to ~$30B by April 2026 — a pace CEO Dario Amodei said outstripped the company's own forecasts by a factor of eight.
| Metric | Value | As of |
|---|---|---|
| ARR run-rate | ~$30B–$43B 🚀 | April–May 2026 |
| YoY revenue growth | ~10× (3 consecutive years) | February 2026 |
| Enterprise share of revenue | ~80% | 2026 |
| $100K+ annual customers | 7× YoY growth | April 2026 |
| $1M+ annual customers | 1,000+ (doubled in <2 mo) | April 2026 |
| Global LLM revenue share | 31.4% (No. 1) | Q1 2026 |
Anthropic led global LLM revenue share in Q1 2026 at 31.4%, narrowly ahead of OpenAI at 29%. As of April 2026, 100,000+ customers run Claude on AWS Bedrock; Anthropic reports revenue from cloud resellers on a gross basis — counting total end-customer spend as revenue and booking partner payouts as expenses. This accounting convention inflates top-line figures vs. net-reporting peers — a key line item for public market investors to scrutinize in the S-1.
The single most important product development in Anthropic's history is also the fastest product to $1B ARR ever built. Claude Code became generally available in May 2025, hit $1B in annualized revenue by November 2025, and reached $2.5B in annualized revenue by February 2026 — more than doubling since the beginning of 2026.
Anthropic gets ~80% of its revenue from enterprises, not consumers. While OpenAI dominates consumer AI with 800M+ weekly ChatGPT users, Anthropic built a more capital-efficient business focused on high-value enterprise contracts that are larger, stickier, and compound over time.
Amazon and Google are the key strategic investors, with combined committed capital exceeding $70 billion across equity stakes and cloud infrastructure obligations — Amazon having deployed $13B in equity to date, while Alphabet currently holds a 14% equity stake contractually capped at 15%.
Claude remains the only frontier AI model available on all three major cloud platforms: Amazon Web Services (Bedrock), Google Cloud (Vertex AI), and Microsoft Azure. That multi-cloud distribution is a structural competitive advantage — no rival model has equivalent cross-platform reach.
Infrastructure expansion includes an agreement with SpaceX to use all compute at the Colossus 1 data center in Memphis (300 MW+, equivalent to 220,000+ NVIDIA GPUs), plus a $50B US infrastructure buildout with Fluidstack with Texas and New York data centers coming online through 2026.
Anthropic projects breaking even by 2028 — two years ahead of OpenAI's 2030 profitability target — a distinction public market investors are likely to price as a premium.
While Anthropic lost ~$5.6B in 2024, the company expects it will achieve $70B in revenue and $17B in positive free cash flow by 2028. Meanwhile, OpenAI projects it will lose $74B in 2028 but will turn a profit by 2030. This is the central public-market premium argument: Anthropic is a better business even if it's the #2 consumer brand.
Bloomberg has reported the company is weighing an IPO as early as October 2026, with Goldman Sachs, JPMorgan, and Morgan Stanley already in early discussions.
| Milestone | Status | Detail |
|---|---|---|
| Counsel engagement | Done | Wilson Sonsini hired to advise on IPO process |
| Banker pitches | Underway | Goldman Sachs, JPMorgan, Morgan Stanley in early discussions |
| Final private round | Board decision | Series H decision expected at May 2026 board meeting |
| S-1 filing | Pending | No EDGAR filing as of May 20, 2026 — most-watched data drop of 2026 |
| Target listing window | Speculated | October 2026 (Bloomberg) |
Anthropic is finding it difficult to resist pressure to raise what could be its final private round before IPO, and is expected to make a definitive decision on the round at a board meeting in May 2026.
| Risk | Severity | Detail |
|---|---|---|
| Gross vs. net revenue 🔴 | High | Cloud reseller revenue counted gross; real net may be ~$22B vs. $30B reported |
| Compute capacity constraint 🔴 | High | Demand outpacing supply; infrastructure strain impacting reliability at peak hours |
| Valuation multiple compression 🟡 | Medium | ~$900B implied vs. $30B ARR = ~30× revenue — extreme even for hypergrowth |
| Open-source commoditization 🟡 | Medium | Meta Llama, DeepSeek, Chinese models compress API pricing structurally |
| Related-party concentration 🟡 | Medium | AWS + Google = >70% of compute + distribution; customer and vendor simultaneously |
| Profitability path dependence | Medium | $70B 2028 revenue target requires continued 3× annual growth — not guaranteed |
Anthropic's monthly active user base is significantly smaller than OpenAI's (134M vs. 900M), but approximately 80% of revenue derives from enterprise customers — compared with roughly 40% for OpenAI. This gives Anthropic structurally higher revenue quality: lower churn, larger contract sizes, and multi-year sticky commitments.
No S-1 filed. Public retail access is currently limited to:
| Vehicle | Ticker | Type | Anthropic Exposure |
|---|---|---|---|
| AI-focused ETF | AGIX | Exchange-listed ETF | ~2.6% weight |
| ARK Venture Fund | ARKVX | Closed-end, private AI basket | ~3.0% weight |
| Destiny Tech100 | DXYZ | Private AI basket | Diversified |
| Amazon (strategic investor) | AMZN | Common stock | Indirect via $13B+ equity stake |
| Alphabet (strategic investor) | GOOGL | Common stock | Indirect via 14% equity stake |
Anthropic is the most compelling pre-IPO technology asset of 2026 — #1 in global LLM revenue share, fastest path to profitability among the hyperscale AI labs, and a genuinely differentiated enterprise revenue model. The primary underwriting risks are gross-vs.-net revenue accounting, extreme capex dependency on AWS/Google, and a valuation multiple that prices in flawless execution through 2028. The S-1 filing, when it hits EDGAR, will be the single most important data drop for sizing both the direct IPO trade and the read-across to AMZN, GOOGL, and the broader AI infrastructure complex.