What is Anthropic known for?

The Claude models, Constitutional AI and interpretability research on what happens inside models. Landmarks: Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback (2022); Constitutional AI: Harmlessness from AI Feedback (2022); Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet (2024); Introducing the Model Context Protocol (2024).

Dario Amodei, Chief Executive Officer and co-founder; Jared Kaplan, Chief Science Officer and co-founder; Rahul Patil, Chief Technology Officer.

What kind of organization is Anthropic?

Startup lab, founded in 2021, based in San Francisco.

Anthropic

Startup lab · founded 2021 · San Francisco, United States · about 2,500 people (Jan 2026)

Website

The Claude models, Constitutional AI and interpretability research on what happens inside models.

Known for

Year	Contribution	Source
2022	Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackarxiv.org	arxiv.org
2022	Constitutional AI: Harmlessness from AI Feedbackarxiv.org	arxiv.org
2024	Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnettransformer-circuits.pub	transformer-circuits.pub
2024	Introducing the Model Context Protocolanthropic.com	anthropic.com

Leaders

On Leaders

Person	Role	Source
Dario AmodeiChief Executive Officer and co-founderanthropic.com	Chief Executive Officer and co-founder · since 2021	anthropic.com
Jared KaplanChief Science Officer and co-founderen.wikipedia.org	Chief Science Officer and co-founder · since 2021	en.wikipedia.org
Rahul PatilChief Technology Officertechcrunch.com	Chief Technology Officer · since Oct 2025	techcrunch.com

Models

On Models

Model	Standing
Claude Fable 5.1Flagship	#1 on the Models podium
Claude Opus 5	#3 on the Models podium
Claude Fable 5	#4 on the Models podium
Claude Opus 5.5	#5 on the Models podium
Claude Opus 4.6	#7 on the Models podium

Papers on Papers

Lab page on Papers

Date	Paper
Jan 2025	Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming
Dec 2024	Alignment faking in large language models
Jan 2024	Sleeper Agents: Training Deceptive LLMs that Persist Through Safety Training
Dec 2022	Constitutional AI: Harmlessness from AI Feedback

Funding and compute

Date	Deal	Source
May 28, 2026	Series H$65B, at a $965B post-money valuation, led by Altimeter, Dragoneer, Greenoaks and Sequoiaanthropic.com	anthropic.com
May 6, 2026	Compute: SpaceXCompute partnershipanthropic.com	anthropic.com
Apr 20, 2026	Compute: AmazonUp to 5 GW of Trainium capacityanthropic.com	anthropic.com
Apr 6, 2026	Compute: Google and BroadcomMultiple gigawatts of next-generation TPU capacity from 2027anthropic.com	anthropic.com
Feb 12, 2026	Series G$30B, at a $380B post-money valuation, led by GIC and Coatueanthropic.com	anthropic.com
Nov 18, 2025	Compute: Microsoft and NVIDIA$30B of Azure compute, up to one gigawatt moreanthropic.com	anthropic.com
Oct 23, 2025	Compute: Google CloudUp to one million TPUs, well over a gigawatt online in 2026anthropic.com	anthropic.com

Links

Where	Link
Website	anthropic.com
Blog	anthropic.com/news
GitHub	anthropics
Hugging Face	anthropic
X	@AnthropicAI
Papers	Lab page on Papers
Models	12 ranked models on Models
Leaders	Leadership on Leaders
Companies	Anthropic across fru.dev

About, sources and history

Type: Startup lab, ai research
Headcount: About 2,500 (Jan 2026), wikidata.org
Sources: wikidata.org Q116758847
Registry: Anthropic on Companies

History

2026-09-24 Added to Labs