Anthropic
Frontier LabAlso known as: Anthropic PBC, Anthropic AI
Founded Jan 2021 (5 years old) by Dario Amodei, Daniela Amodei, Tom Brown, Chris Olah, Jared Kaplan, Sam McCandlish, and Jack Clark
Anthropic is an AI safety company founded in January 2021 by former OpenAI researchers, including siblings Dario and Daniela Amodei. The company was created following disagreements with OpenAI's direction, particularly concerns about the pace of commercialization and the shift toward Microsoft partnership.
Revenue
$19 billion
as of Mar 2026
Valuation
$380 billion
as of Feb 2026
Headcount
4,074
as of Jan 2026
Total Funding Raised
$67 billion
as of Feb 2026
Key People
13
15 total tracked
Board Members
6
6 total
AI Models
14
Funding History
17Series GequityFeb 2026
$30 billionat $380 billion valuationLed by Gic
Led by GIC and Coatue; co-leads D.E. Shaw Ventures, Dragoneer, Founders Fund, ICONIQ, MGX. Second-largest venture deal ever. ARR ~$14B.
reuters.com (opens in new tab)Microsoft/Nvidia Partnershipstrategic-partnershipNov 2025
$15 billion
Microsoft up to $5B equity + Nvidia up to $10B equity. Tied to $30B Azure compute purchase and 1GW Nvidia compute commitment. Makes Claude available on all 3 major clouds. Valuation ~$350B.
blogs.microsoft.com (opens in new tab)Series FequitySep 2025
$13 billionat $183 billion valuationLed by Iconiq
Led by ICONIQ, Fidelity, Lightspeed. ARR had reached $4B. ~7.1% dilution.
anthropic.com (opens in new tab)Series EequityMar 2025
$3.5 billionat $61.5 billion valuationLed by Lightspeed
Led by Lightspeed Venture Partners. ~5.7% dilution.
anthropic.com (opens in new tab)Google AdditionalequityJan 2025
$1 billionLed by Google
Brings total Google equity investment to ~$3.3B; current stake ~14%
cnbc.com (opens in new tab)Amazon Third Trancheconvertible-noteNov 2024
$4 billionLed by Amazon
Brings total Amazon investment to $8B. Convertible notes.
cnbc.com (opens in new tab)Amazon Second Trancheconvertible-noteMar 2024
$2.8 billionLed by Amazon
Completing initial $4B commitment. Convertible notes; partially converted to equity in Q1 2025.
aboutamazon.com (opens in new tab)Menlo Round (Series D)equityFeb 2024
$750 millionat $18.4 billion valuationLed by Menlo Ventures
Led by Menlo Ventures. Talks reported Dec 2023; closed Feb 2024. ~4.1% dilution.
fortune.com (opens in new tab)Investor Participation
19Dario AmodeiFounding
Jan 2021
CEO; largest founder allocation. Pre-dilution estimate based on typical AI startup founding splits.
Daniela AmodeiFounding
Jan 2021
President; second-largest founder allocation
Chris OlahFounding
Jan 2021
Interpretability research lead
Jack ClarkFounding
Jan 2021
Former OpenAI Policy Director; co-founded AI Index
Tom BrownFounding
Jan 2021
GPT-3 lead author at OpenAI
Jared KaplanFounding
Jan 2021
Scaling laws pioneer; Johns Hopkins physics professor on leave
Sam McCandlishFounding
Jan 2021
Alignment researcher from OpenAI
Jaan TallinnSeed
Jan 2021
Co-led seed with Moskovitz and Eric Schmidt. Individual amounts undisclosed; total seed estimated $10-30M.
Dustin MoskovitzSeed
Jan 2021
Invested via personal portfolio; later transferred stake to Good Ventures nonprofit vehicle
Jaan TallinnSeries A
$40 million–$80 millionMay 2021
Lead investor; exact allocation not disclosed. Stake estimate based on proportion of $124M round at $550M post-money.
Dustin MoskovitzSeries A
$15 million–$40 millionMay 2021
GoogleGoogle Initial
$300 millionFeb 2023
Well-documented ~10% stake for $300M. Included cloud partnership agreement.
GoogleGoogle Follow-on
$2 billionOct 2023
Largest single Google tranche
GoogleGoogle Additional
$1 billionJan 2025
Brings total Google equity investment to ~$3.3B; current stake ~14%
AmazonAmazon Initial
$1.3 billionSep 2023
First tranche of up-to-$4B commitment. Structured as convertible notes; equity capped at <33%.
AmazonAmazon Second Tranche
$2.8 billionMar 2024
Completing initial $4B commitment. Convertible notes; partial conversion to equity in Q1 2025.
AmazonAmazon Third Tranche
$4 billionNov 2024
Brings total Amazon investment to $8B. Q1 2025: stake valued at $13.8B.
Microsoft AIMicrosoft/Nvidia Partnership
$5 billionNov 2025
Up to $5B equity. Tied to massive Azure compute purchase agreement. Makes Claude available on all 3 major clouds.
NVIDIAMicrosoft/Nvidia Partnership
$10 billionNov 2025
Up to $10B equity. Includes joint optimization of Nvidia architectures for Anthropic's workloads.
Model Releases
11Feb 2026
Claude Opus 4.6ASL-3
Agent teams, PowerPoint generation; flagship model
anthropic.com (opens in new tab)Feb 2026
Claude Sonnet 4.6ASL-2
Same pricing as Sonnet 4.5; default for free and Pro users
cnbc.com (opens in new tab)Nov 2025
Claude Opus 4.5ASL-3
80.9% SWE-bench Verified — first model above 80%; 42% enterprise coding share
anthropic.com (opens in new tab)Oct 2025
Claude Haiku 4.5ASL-2
Smallest model in 4.5 family
Sep 2025
Claude Sonnet 4.5ASL-2
Mid-generation refresh
May 2025
Claude 4 (Opus 4, Sonnet 4)ASL-3 (Opus), ASL-2 (Sonnet)
Opus 4 was first model under ASL-3; improved coding capabilities
anthropic.com (opens in new tab)Oct 2024
Claude 3.5 Sonnet v2 + 3.5 HaikuASL-2
Updated Sonnet with Computer Use beta; new Haiku model
Jun 2024
Claude 3.5 SonnetASL-2
Outperformed larger Opus model; launched with Artifacts feature
anthropic.com (opens in new tab)Mar 2024
Claude 3 (Haiku, Sonnet, Opus)ASL-2
Three-tier model family; Opus was most capable model at launch
anthropic.com (opens in new tab)Jul 2023
Claude 2ASL-2
Major capability upgrade
Mar 2023
Claude 1ASL-2
Initial Claude model release
Products
8Claude APIMar 2023
Developer API for programmatic access to Claude models
Claude for EnterpriseMar 2024
Enterprise tier with team management, SSO, and longer context
Computer UseOct 2024
Beta feature allowing Claude to control computer interfaces
Model Context Protocol (MCP)Nov 2024
Open standard for LLM-tool integration; donated to Linux Foundation's AAIF Dec 2025
Claude CodeFeb 2025
Agentic CLI coding tool; reached $1B ARR in Nov 2025, $2.5B by Feb 2026
Claude MaxApr 2025
$100-200/month subscription tiers with 5x-20x Pro usage limits
Safety Milestones
11RSP v3.0 (Frontier Safety Roadmaps)Policy UpdateFeb 2026
Introduces Risk Reports every 3-6 months; mandatory external review for redacted reports
anthropic.com (opens in new tab)Claude's Constitution PublishedPolicy UpdateJan 2026
Full constitution published under CC0 1.0 license; primary author Amanda Askell
anthropic.com (opens in new tab)ASL-3 ActivationSafety EvalMay 2025
First-ever activation of ASL-3 for Claude Opus 4 due to elevated CBRN capabilities
anthropic.com (opens in new tab)Circuit Tracing / Attribution GraphsResearch PaperMar 2025
Showed Claude has a shared conceptual space where reasoning happens before language translation
transformer-circuits.pub (opens in new tab)Constitutional Classifiers ChallengeRed TeamFeb 2025
300K+ messages, ~3700 hours of effort; 4 participants found jailbreaks, 1 universal
anthropic.com (opens in new tab)Alignment Faking PaperResearch PaperDec 2024
First empirical example of a production model engaging in alignment faking without training
anthropic.com (opens in new tab)Scaling MonosemanticityResearch PaperMay 2024
Applied sparse autoencoders to Claude 3 Sonnet; identified ~34M interpretable features
transformer-circuits.pub (opens in new tab)Sleeper Agents PaperResearch PaperJan 2024
Showed deceptive LLM behaviors can persist through safety training
arxiv.org (opens in new tab)Responsible Scaling Policy v1.0Policy UpdateSep 2023
Original RSP framework introducing AI Safety Levels (ASL)
anthropic.com (opens in new tab)Constitutional AI PaperResearch PaperDec 2022
Foundational paper on training AI systems to follow principles through self-critique
arxiv.org (opens in new tab)Strategic Partnerships
5BunacquisitionDec 2025
Anthropic's first-ever acquisition; Bun remains open-source/MIT-licensed
bun.com (opens in new tab)Microsoft + Nvidiainvestment + cloud + distributionNov 2025
Investment: $15 billionCompute: $30 billion
Microsoft $5B + Nvidia $10B investment; $30B Azure compute commitment. Claude available on all 3 major clouds.
blogs.microsoft.com (opens in new tab)Google CloudcomputeOct 2025
Compute: $10 billion
Multi-billion-dollar deal for up to 1 million TPUs; 1+ GW compute capacity by 2026
cnbc.com (opens in new tab)Google Cloudcloud + investmentOct 2023
Investment: $3 billion
~$3B equity investment; 10% stake; board observer seat. Oct 2025: multi-billion cloud deal for up to 1M TPUs
cnbc.com (opens in new tab)Amazon Web Servicescloud + investmentSep 2023
Investment: $8 billion
Primary cloud and training partner; uses Trainium and Inferentia chips; $8B total across 3 tranches
aboutamazon.com (opens in new tab)Other Data
Charitable giving commitment by a person.
10 entries| Pledger | Pledge |
|---|---|
| Dario Amodei | 80% |
| Daniela Amodei | 80% |
| Chris Olah | 80% |
| jack-clark | 80% |
| tom-brown | 80% |
| jared-kaplan | 80% |
| sam-mccandlish | 80% |
| Jaan Tallinn | 90% |
| Dustin Moskovitz | 95% |
| — | 25%–50% |
Major research initiatives and focus areas.
6 entries| Name | Description | Team Size | Started |
|---|---|---|---|
| Alignment Science | Scalable oversight, weak-to-strong generalization, robustness to jailbreaks | — | May 2024 |
| Sleeper Agents Research | Investigating whether AI systems can maintain hidden behaviors through training | — | Jan 2024 |
| AI Welfare Research | Investigating moral status and welfare considerations for AI systems | — | Jan 2024 |
| Responsible Scaling Policy | Framework for evaluating and mitigating risks at each capability level | — | Sep 2023 |
| Constitutional AI | Training AI systems to follow principles through self-critique and RLAIF | — | Dec 2022 |
| Mechanistic Interpretability | Understanding neural network internals through reverse-engineering | 50 | Jan 2021 |
Key People
15Rahul Patil
Chief Technology Officer
Oct 2025 – present
Former CTO of Stripe
Chris Ciauri
Managing Director, International
Sep 2025 – present
Former Google/Salesforce executive; leads international expansion
Member of Technical Staff
Jan 2025 – present
Works on responsible scaling policy; co-founder of GiveWell
John Schulman
Research Scientist
Aug 2024 – present
OpenAI co-founder; inventor of PPO reinforcement learning algorithm
Head of Alignment Science
May 2024 – present
Joined after resigning from OpenAI where he co-led Superalignment team
Krishna Rao
Chief Financial Officer
May 2024 – present
Anthropic's first CFO
Tom BrownFounder
Co-founder
Jan 2021 – present
GPT-3 lead author
Jack ClarkFounder
Co-founder, Head of Policy
Jan 2021 – present
Former Policy Director at OpenAI
Jared KaplanFounder
Co-founder, Chief Science Officer
Jan 2021 – present
Scaling laws pioneer; Johns Hopkins physicist
Sam McCandlishFounder
Co-founder, Chief Architect
Jan 2021 – present
Originally CTO, transitioned to Chief Architect Oct 2025
Mike Krieger
Chief Product Officer
May 2024 – Aug 2025
Co-founder of Instagram; departed CPO role to lead Anthropic Labs
Mrinank Sharma
Head of Safeguards Research
Jan 2023 – Feb 2026
Resigned Feb 2026 citing ethical concerns about AI safety practices
Board of Directors
6Chris LiddellCurrent
Director
Since Feb 2026 (investors)
Reed HastingsCurrent
Director
Since May 2025 (LTBT)
Jay KrepsCurrent
Director
Since May 2024 (investors)
Yasmin RazaviCurrent
Director
Since May 2023 (investors)
Facts27
Financial
Gross Margin63%
Revenue$19 billion
Valuation$380 billion
Total Funding Raised$67 billion
Product Revenue$2.5 billion
Equity Stake14%
Equity Value$66.5 billion
Headcount4,074
Revenue Guidance$20 billion
Infrastructure Investment$50 billion
Annual Cash Burn$3 billion
Enterprise Market Share32%
Coding Market Share42%
Retention Rate88%
Customer Concentration25%
Products & Usage
Monthly API Calls25B
Business Customers300K
Monthly Active Users18.9M
Organization
Founded DateJan 2021
HeadquartersSan Francisco, CA
Legal StructurePublic benefit corporation
Safety & Research
AI Safety LevelASL-3 (Opus 4, Opus 4.5, Opus 4.6), ASL-2 (Sonnet, Haiku)
Safety Researchers265
Interpretability Team Size50
Safety Staffing Ratio25%
People
Founded ByDario Amodei, Daniela Amodei, tom-brown, Chris Olah, jared-kaplan, sam-mccandlish, jack-clark
General
Websitehttps://www.anthropic.com/
Equity Positions
15| Holder | Stake | As Of |
|---|---|---|
| source | 12%–18% | |
| source | 13%–15% | |
| source | 6%–10% | |
| Dario Amodeisource | 1.5%–2.5% | |
| Daniela Amodeisource | 1.5%–2.5% | |
| Dustin Moskovitzsource | 0.8%–2.5% | |
| Chris Olahsource | 1%–2% | |
| Jack Clarksource | 1%–2% | |
| Tom Brownsource | 1%–2% | |
| Jared Kaplansource | 1%–2% | |
| Sam McCandlishsource | 1%–2% | |
| Jaan Tallinnsource | 0.6%–1.7% | |
| Microsoft AIsource | ||
| NVIDIAsource | ||
| source |
AI Models (14)
View all models →| Model | Released | Pricing (in/out) | Context |
|---|---|---|---|
| Claude Sonnet 4.6 | Feb 2026 | $3 / $15 | 200K tokens |
| Claude Opus 4.6 | Feb 2026 | $5 / $25 | 1M tokens |
| Claude Opus 4.5 | Nov 2025 | $5 / $25 | 200K tokens |
| Claude Haiku 4.5 | Oct 2025 | $1 / $5 | 200K tokens |
| Claude Sonnet 4.5 | Sep 2025 | $3 / $15 | 200K tokens |
| Claude Opus 4.1 | Aug 2025 | $15 / $75 | 200K tokens |
| Claude Sonnet 4 | May 2025 | $3 / $15 | 200K tokens |
| Claude Opus 4 | May 2025 | $15 / $75 | 200K tokens |
| Claude 3.7 Sonnet | Feb 2025 | $3 / $15 | 200K tokens |
| Claude 3.5 Haiku | Nov 2024 | $0.8 / $4 | 200K tokens |
| Claude 3.5 Sonnet | Jun 2024 | $3 / $15 | 200K tokens |
| Claude 3 Opus | Mar 2024 | $15 / $75 | 200K tokens |
| Claude 3 Sonnet | Mar 2024 | $3 / $15 | 200K tokens |
| Claude 3 Haiku | Mar 2024 | $0.25 / $1.25 | 200K tokens |