This is not a rumor. Both numbers come from official company announcements. And the story behind how it happened is more interesting than the number itself.
Quick Summary
| Metric | Figure | Source |
|---|---|---|
| Anthropic valuation | $965 billion | Anthropic official, May 28, 2026 |
| Anthropic Series H raise | $65 billion | Anthropic official |
| Anthropic revenue run rate | $47 billion (annualized, not audited) | Anthropic CFO statement |
| OpenAI valuation | $852 billion | OpenAI official, March 31, 2026 |
| OpenAI monthly revenue | $2 billion/month | OpenAI official |
| Claude enterprise win rate vs ChatGPT | ~70% | Ramp AI Index, 50K+ companies |
| Google investment in Anthropic | $40 billion | Confirmed, April 2026 |
| Amazon investment in Anthropic | Up to $28 billion | Confirmed, April 2026 |
| Anthropic IPO target | October 2026 | Reported, not officially confirmed |
| OpenAI IPO filing | May 22, 2026 | Confirmed — CNBC |
Sourcing note: Every figure is from official announcements or Bloomberg, CNBC, or TechCrunch reporting. Where numbers are annualized run rates rather than audited revenue, that is labeled clearly.
What Actually Happened
On May 28, 2026, Anthropic published an official announcement confirming a $65 billion Series H funding round at a $965 billion post-money valuation. The round was led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital. Bloomberg, CNBC, and TechCrunch each independently confirmed the figures the same day.
In the same announcement, Anthropic’s CFO Krishna Rao stated:
“Since our Series G in February, adoption has continued to grow across global enterprise customers, and our run-rate revenue crossed $47 billion earlier this month.”
That $47 billion is an annualized run rate — a projection based on recent monthly performance, not a GAAP-audited annual total. That distinction matters and is addressed in detail below.
For comparison, OpenAI’s $852 billion valuation was confirmed when OpenAI officially closed a $122 billion funding round on March 31, 2026, led by Amazon, Nvidia, and SoftBank. Both numbers are real. Anthropic’s is now larger.
Who Is Anthropic?
Most people know ChatGPT. Far fewer know Claude by name — at least in consumer settings. But in enterprise environments, the story is very different, and that gap is the core of this whole story.
The Founding
Anthropic was founded in 2021 by Dario Amodei (CEO) and Daniela Amodei (President), along with six other researchers — all formerly of OpenAI. Dario had been VP of Research at OpenAI. Daniela had been VP of Safety and Policy.
They did not leave over money or ego. They left because they believed AI safety was being underweighted relative to how fast capabilities were advancing. In their view, the right response was not to argue from within OpenAI but to build a new institution with different incentives and legal structure from the ground up.
Anthropic was incorporated as a public benefit corporation — a structure that lets the board legally prioritize the mission of beneficial AI over short-term profit. That choice, which looked idealistic in 2021, became a competitive advantage by 2026, particularly with enterprise buyers who need AI vendors they can defend to their legal and compliance teams.
Constitutional AI
In 2022, Anthropic published its Constitutional AI (CAI) methodology — a training approach that teaches models to evaluate their own outputs against a written set of principles. The result is a model that is more predictable, more consistent, and less likely to produce unexpected outputs at the edge cases that matter most in professional settings. This is the technical foundation of Claude’s enterprise reputation.
The Funding Timeline
Every round below is confirmed from official sources:
| Round | Date | Amount | Post-Money Valuation |
|---|---|---|---|
| Series A | 2022 | $580M | ~$1.3B |
| Series B | 2023 | $1.25B | ~$4.1B |
| Series C | May 2023 | $450M | ~$5B |
| Series D | Late 2023 | $1.25B | ~$15B |
| Series E | 2024 | $2.75B | ~$18.4B |
| Series F | Sept 2025 | $13B | $183B |
| Series G | Feb 2026 | $30B | $380B |
| Series H | May 2026 | $65B | $965B |
Total raised: approximately $132 billion across 18 rounds (Tracxn, May 2026).
The 15-fold valuation increase from $61.5 billion in March 2025 to $965 billion in May 2026 — in 14 months — has no clear precedent in private market history.
Understanding the Revenue Numbers
This section matters. Anthropic’s revenue figures have been widely quoted and widely misunderstood.
What a run rate actually means
A run rate is an annualized estimate based on recent monthly performance. If Anthropic earned approximately $3.9 billion in a given month, the annualized run rate is approximately $47 billion. It does not mean Anthropic has collected $47 billion in the past year. It means: if the current pace holds for twelve months, the annual total would be $47 billion.
Run rates are useful directional signals. They are not substitutes for audited revenue. The GAAP-audited figures will only become public when Anthropic files its S-1 IPO prospectus. Until then, all revenue figures should be treated as directional.
The trajectory (all verified)
| Period | Figure | Type |
|---|---|---|
| Full year 2024 | ~$1B | Annual revenue |
| Full year 2025 | ~$9B | Annual revenue — Anthropic Series G official |
| February 2026 | $14B | Annualized run rate — Anthropic official |
| April 2026 | $30B+ | Annualized run rate — Anthropic confirmed |
| May 2026 | $47B | Annualized run rate — Anthropic CFO official |
Going from $1 billion in annual revenue to a $47 billion run rate in approximately 18 months is one of the fastest revenue growth arcs in enterprise software history, by any reasonable measure.
OpenAI for comparison: OpenAI confirmed in its March 31, 2026 official announcement that it is generating $2 billion in monthly revenue, implying approximately $24 billion annualized. On a run-rate basis, Anthropic has now surpassed OpenAI — a reversal that would have seemed implausible twelve months ago.
Why Revenue Grew So Fast: Claude Code
The single biggest driver of Anthropic’s revenue explosion is a product that most non-developers have never heard of: Claude Code.
Claude Code is Anthropic’s AI coding assistant, generally available since May 2025. The verified trajectory:
- November 2025: $1 billion annualized run rate
- February 2026: $2.5 billion annualized run rate (Anthropic official, Series G)
- May 2026: Estimated to represent roughly one-third of total Anthropic revenue
Enterprise customers confirmed by Sacra (May 2026) to use Claude Code in production include Netflix, Spotify, KPMG, L’Oréal, and Salesforce.
The product became a cultural phenomenon among software engineers. Developers started describing themselves as “Claude-pilled” — deeply integrated with Claude to the point where switching tools felt actively inefficient. The technical reason: Claude can process an entire large codebase in a single context window, understand cross-file dependencies, and suggest changes that account for the full system rather than isolated snippets.
Claude Code competes with GitHub Copilot, Cursor, and OpenAI’s Codex. Its edge is not price — it is reliability, context window size, and instruction-following consistency in production environments.
Claude vs. ChatGPT: A Fact-Based Comparison
Neither model is universally better. The correct choice depends on your use case. Here is what the verified data shows.
Where Claude leads
Coding accuracy: An independent 30-day test by Ryz Labs found Claude achieved approximately 95% functional accuracy on coding tasks versus approximately 85% for ChatGPT.
Context window: Claude’s default context window is 200,000 tokens versus ChatGPT’s 128,000 tokens. Claude Opus 4.7 (April 2026) introduced a 1 million token window in production. Claude shows less than 5% accuracy degradation across its full 200K range.
Reasoning: On GPQA Diamond (PhD-level science questions), Claude scores 91.3%. Professional users in writing-intensive fields consistently report Claude follows complex, nuanced instructions more faithfully.
Enterprise win rate: The Ramp March 2026 AI Index, tracking real corporate spending across more than 50,000 businesses, found Claude wins approximately 70% of head-to-head enterprise AI purchasing decisions against ChatGPT. Anthropic holds approximately 29% of the enterprise AI assistant market (DemandSage, 2026), and more than 1,000 enterprise customers now spend over $1 million annually on Claude.
Where ChatGPT leads
ChatGPT Plus ($20/month) includes integrated image generation, voice, and video — a broader multimodal ecosystem Claude Pro does not match at the same price. ChatGPT holds approximately 68% of global AI chatbot web traffic versus Claude’s approximately 4.5% (SimilarWeb, January 2026). OpenAI confirmed 900 million weekly active users as of March 2026 — a consumer user base Anthropic does not approach.
API pricing
| Model | Input (per 1M tokens) | Output (per 1M tokens) |
|---|---|---|
| Claude Opus 4.6 | $5 | $25 |
| Claude Sonnet 4.6 | $3 | $15 |
| Claude Haiku 4.5 | $1 | $5 |
Batch API discounts reduce costs by 50%. Prompt caching reduces repeat input costs by up to 90%.
The honest summary: Consumer AI belongs to ChatGPT. Enterprise AI is increasingly Claude’s territory. Both are excellent tools. The choice should come down to your actual workflow — not brand recognition.
Why Google and Amazon Are Investing in Their Competitor
This is the most counterintuitive part of the Anthropic story.
Google’s $40 billion
In April 2026, Google and Anthropic officially confirmed Google is investing $40 billion in Anthropic — despite Google competing with Claude through its own Gemini AI. The investment comes alongside an expanded cloud partnership where Anthropic will significantly scale its use of Google TPUs.
The logic is cloud economics, not brand loyalty. Every Claude workload running on Google Cloud generates infrastructure revenue for Google regardless of which AI model wins the product competition. Independent analysis estimates cloud providers capture 16–32% of AI startup revenue through infrastructure relationships. Google wins either way.
Amazon’s $28 billion
Amazon has committed up to $28 billion total in Anthropic, including $5 billion in the Series H. Anthropic has committed to spend over $100 billion on AWS over the next ten years. Amazon’s original $8 billion stake is now worth more than $70 billion according to Amazon’s own Q1 2026 earnings — representing more than half of Amazon’s pre-tax income for the quarter.
That is not a typo. One AI investment contributed more to Amazon’s quarterly pre-tax income than most of its other business lines combined.
The multi-cloud advantage
Claude runs across Google TPUs, Amazon Trainium, and Nvidia GPUs simultaneously. This multi-cloud architecture gives Anthropic infrastructure redundancy and negotiating leverage that few companies its size could achieve.
OpenAI’s Position: Still Dominant, But Under Pressure
OpenAI is not losing. This needs to be stated plainly. ChatGPT has 900 million weekly active users. OpenAI generates $2 billion in monthly revenue. The ChatGPT brand is the most recognized AI product in the world. None of that has changed.
But there are structural pressures worth understanding.
Profitability: OpenAI projects reaching profitability no earlier than 2029. Projected losses for 2026 are estimated at up to $14 billion (The Information, cited by multiple outlets). Infrastructure spending commitments are enormous.
Microsoft dependency: Microsoft takes 75% of all OpenAI profits until its investment is recouped, and 49% thereafter until reaching $92 billion. That structure constrains OpenAI’s financial independence regardless of revenue growth.
IPO: OpenAI filed its IPO prospectus confidentially with the SEC on May 22, 2026 (CNBC confirmed), targeting a public listing as early as September 2026. Note: $35 billion of Amazon’s $50 billion OpenAI investment is contingent on OpenAI completing an IPO by 2028 or achieving AGI.
Market share: OpenAI’s global AI chatbot web traffic share declined from approximately 87% to 68% over twelve months (SimilarWeb, January 2026). The market has fragmented across Gemini, Claude, Grok, and Perplexity.
The Broader AI Race: Where Everyone Stands
The AI industry has settled into three distinct strategic positions:
Consumer scale — OpenAI/ChatGPT: Largest user base, broad multimodal product, deep Microsoft distribution. Challenge: proving profitability before the IPO.
Ecosystem distribution — Google/Gemini: Gemini’s traffic share grew from 5.4% to 18.2% in twelve months. Google’s advantage is distribution — Gemini appears inside Search, Gmail, Workspace, and Android without requiring users to change habits.
Enterprise precision — Anthropic/Claude: Small consumer footprint, dominant enterprise win rate, coding-first positioning, safety-first brand reputation in risk-sensitive industries.
Beyond these three: Meta’s Llama 4 threatens to commoditize the API market through open-source self-hosting. xAI’s Grok 4.3 offers API pricing approximately 75% below frontier leaders, creating meaningful price pressure.
The Real Risks: What Could Go Wrong
1. Run rate vs. audited revenue
The $47 billion figure is a company-disclosed run rate, not GAAP-audited revenue. When Anthropic files its S-1, audited figures will be disclosed for the first time. If enterprise contracts include front-loaded payments, deferred revenue, or usage-variable terms, the GAAP figures could diverge meaningfully from the run-rate headline.
2. Government conflict
In February 2026, President Trump directed federal agencies to immediately cease using Anthropic’s technology, and the Department of Defense designated Anthropic a “supply chain risk to national security” (confirmed by the Congressional Research Service, April 2026). The conflict followed Anthropic’s refusal to accept certain military applications — specifically AI-controlled weapons and mass domestic surveillance. A court subsequently issued a temporary block on the designation. Government-adjacent enterprise market access remains a risk factor.
3. Infrastructure dependency
Anthropic has committed over $100 billion in AWS spending over ten years. That creates capacity certainty but also strategic dependency on a partner whose competitive interests do not always align with Anthropic’s.
4. Competition is closing the gap
Every Claude advantage — coding accuracy, context window, instruction-following — is being actively targeted by OpenAI (Codex), Google (Gemini), Meta (Llama), and xAI (Grok). A 70% enterprise win rate today is not a structural guarantee tomorrow.
5. Valuation multiples
Anthropic’s $965 billion valuation represents approximately 107x its confirmed 2025 annual revenue of $9 billion. Multiples at this level price in extraordinary sustained growth. They leave significant room for disappointment if growth moderates or the IPO process reveals structural weaknesses.
What Happens Next: Three Scenarios
Most likely: structured competition
Both Anthropic and OpenAI complete IPOs in 2026 or early 2027. Public market scrutiny moderates both valuations. The AI market stabilizes into a durable dual-platform structure — ChatGPT for consumer and general productivity, Claude for enterprise and coding workflows. Both companies generate substantial revenue. The global AI market, projected at $371 billion in 2025 and $2.4 trillion by 2032 (MarketsandMarkets), is large enough for multiple winners.
Best case for Anthropic
Claude Code’s workflow integration deepens to the point where enterprise switching costs become prohibitive. The IPO prices at or above $965 billion. Anthropic reaches operating profitability in Q2 2026 (sources cited by CNBC suggest approximately $559 million operating profit may be on track). The company demonstrates a credible path to sustained profitability before the IPO, strengthening its public-market positioning significantly.
Worst case for Anthropic
The GAAP revenue disclosed in the S-1 is materially below the $47 billion run rate, triggering a valuation reset. A major enterprise customer exits publicly following a high-profile AI failure. OpenAI closes the coding performance gap, eliminating Claude Code’s core differentiator. The government conflict expands. The IPO prices below private market expectations.
The Founding Bet, Revisited
In December 2020, Dario Amodei and fourteen colleagues left OpenAI. Not because the company was failing — it was succeeding. They left because they believed it was succeeding in the wrong direction: building capabilities faster than the ability to make those capabilities safe, controllable, and genuinely beneficial.
They bet that an AI company built around reliability and safety could be commercially competitive. That enterprises, given a real choice, would eventually prefer AI they could trust over AI that was merely fast.
As of May 28, 2026, the private market has priced that bet at $965 billion.
That does not answer every question about Anthropic’s future. Valuation is not profit. Run rate is not audited revenue. Win rates are snapshots. The IPO will reveal things private markets have not yet demanded.
But the data shows something worth noting: the careful, reliability-focused approach to building AI — the one that looked like the slower path in 2021 — has produced a company now valued higher than the one that defined the category.
10 Key Takeaways
- Anthropic’s $965B valuation is officially confirmed by Anthropic, Bloomberg, CNBC, and TechCrunch — May 28, 2026.
- OpenAI’s $852B valuation is also officially confirmed — March 31, 2026. Both are real.
- The $47B revenue figure is an annualized run rate, not audited revenue — the GAAP equivalent will come with the S-1 filing.
- Claude Code drove the growth — from $0 to $2.5B+ run rate in under twelve months from launch.
- Claude wins ~70% of enterprise head-to-head decisions against ChatGPT per Ramp’s 50,000-company dataset. ChatGPT dominates consumer at 68% web traffic share.
- Google invested $40B in Anthropic despite competing with it through Gemini — cloud infrastructure economics explain the logic.
- Amazon’s $8B stake is now worth $70B+, representing more than half of Amazon’s Q1 2026 pre-tax income.
- Both companies are racing toward IPOs in 2026 — OpenAI targeting September (filing confirmed), Anthropic targeting October (reported, not confirmed).
- Real risks exist — run-rate vs. GAAP gap, profitability timeline, infrastructure dependency, government conflict, and intensifying competition.
- The AI market has structurally split — consumer (ChatGPT), enterprise (Claude), ecosystem (Gemini). No single company dominates all three.
Frequently Asked Questions
Is Anthropic officially worth more than OpenAI? Yes, by most recent confirmed private valuation. Anthropic’s Series H closed at $965 billion post-money on May 28, 2026. OpenAI’s most recent confirmed valuation is $852 billion from March 31, 2026. Both are private valuations, not public market prices.
Is Claude better than ChatGPT? Neither dominates every category. Claude leads on coding accuracy (~95% vs ~85%), long-document analysis, and enterprise win rate (70%). ChatGPT leads on consumer user volume (900M+ weekly users), multimodal breadth (images, voice, video), and raw web traffic (68%). Choose based on your specific workflow.
What is Claude Code? Claude Code is Anthropic’s AI coding assistant, generally available since May 2025. It reached $2.5B in annualized run-rate revenue by February 2026 (Anthropic official). Enterprise customers include Netflix, Spotify, KPMG, L’Oréal, and Salesforce. It is the primary engine behind Anthropic’s revenue acceleration.
Is the $47 billion revenue figure real? It is real as a company-disclosed annualized run rate, stated by Anthropic’s CFO in the official Series H announcement. It is not GAAP-audited annual revenue. The audited figures will be disclosed when Anthropic files its S-1 IPO prospectus.
Why did Google invest $40 billion in a competitor? Cloud infrastructure economics: every Claude workload on Google Cloud generates infrastructure revenue for Google regardless of which AI model wins the product competition. Cloud providers capture an estimated 16–32% of AI startup revenue through infrastructure relationships.
When is Anthropic going public? Anthropic is targeting an IPO as early as October 2026, with Goldman Sachs, JPMorgan, and Morgan Stanley in early discussions as underwriters, and Wilson Sonsini retained as legal counsel (Bloomberg, March 2026). No S-1 has been publicly filed as of May 31, 2026.
When is OpenAI going public? OpenAI filed its IPO prospectus confidentially with the SEC on May 22, 2026, with Goldman Sachs and Morgan Stanley leading, targeting a public listing as early as September 2026 (CNBC confirmed).
What are the biggest risks for Anthropic? The five key risks are: (1) the gap between run-rate and GAAP audited revenue, (2) ongoing profitability timeline pressures, (3) strategic dependency on Amazon and Google infrastructure, (4) history of federal government conflict, and (5) intensifying competition from OpenAI, Google, Meta, and xAI.
Sources
- Bloomberg — “Anthropic Raises at $965 Billion Valuation” — May 28, 2026
- CNBC — “Anthropic tops OpenAI as most valuable AI startup” — May 28, 2026
- TechCrunch — “Anthropic raises $65 billion” — May 28, 2026
- Fortune — “Half of Google’s and Amazon’s AI profits came from Anthropic” — April 30, 2026
Last updated: May 31, 2026. This article will be updated when either company files a public S-1 prospectus, which will provide the first GAAP-audited financial data and materially update the revenue picture described here.
