# Inside Anthropic

The company behind Claude, researched with Autumn (https://www.autumn.ai) from public sources. Updated 2026-10-18. Interactive version: https://www.autumn.ai/resources/anthropic

- Founded: 2021, by eight former OpenAI staff
- Latest valuation: $965B post-money (Series H, May 2026) [1](https://www.anthropic.com/news/series-h)
- Revenue run-rate: $65B (Jul 2026) [1](https://www.cnbc.com/2026/08/17/anthropic-says-annualized-revenue-climbed-to-65-billion-in-july.html)
- IPO: confidentially filed, Reuters reported on June 1, 2026; no public S-1 as of 2026-10-18. Details: https://www.autumn.ai/resources/anthropic/ipo

## Co-founders

- [Dario Amodei](https://www.autumn.ai/resources/anthropic/people/dario-amodei), CEO: Trained as a biophysicist, worked a stint as a geophysicist, then led research at OpenAI before leaving to start Anthropic.
- [Daniela Amodei](https://www.autumn.ai/resources/anthropic/people/daniela-amodei), President: An English literature major who went from congressional campaigns to Stripe and OpenAI. She now runs Anthropic day to day and chairs its board.
- [Jared Kaplan](https://www.autumn.ai/resources/anthropic/people/jared-kaplan), Chief Science Officer: A theoretical physicist at Johns Hopkins who co-wrote the 2020 scaling-laws paper while consulting for OpenAI.
- [Jack Clark](https://www.autumn.ai/resources/anthropic/people/jack-clark), Head of Public Benefit: A tech journalist at Bloomberg and The Register before OpenAI. He still writes the Import AI newsletter he started in 2016.
- [Chris Olah](https://www.autumn.ai/resources/anthropic/people/chris-olah), Interpretability lead: Left the University of Toronto after one year, without finishing a degree, and became one of the founders of interpretability research.
- [Ben Mann](https://www.autumn.ai/resources/anthropic/people/ben-mann), Labs co-lead: One of the architects of GPT-3 at OpenAI. His own site calls him a software engineer, tinkerer and aspiring mad scientist.
- [Sam McCandlish](https://www.autumn.ai/resources/anthropic/people/sam-mccandlish), Chief Architect: A theoretical physicist who wrote his Stanford PhD on quantum gravity, then co-wrote GPT-3 and the scaling-laws paper.
- [Tom Brown](https://www.autumn.ai/resources/anthropic/people/tom-brown), Chief Compute Officer: A startup engineer who taught himself AI research and went on to lead the engineering of GPT-3.

## Timeline

- Jan 2021: Founded [1](https://en.wikipedia.org/wiki/Anthropic)
- May 2021: Series A: raised $124M, led by Jaan Tallinn [1](https://www.anthropic.com/news/anthropic-raises-124-million-to-build-more-reliable-general-ai-systems)
- Apr 2022: Series B: raised $580M, led by Sam Bankman-Fried [1](https://www.anthropic.com/news/anthropic-raises-series-b-to-build-safe-reliable-ai)
- Dec 2022: Constitutional AI paper [1](https://www.anthropic.com/research/constitutional-ai-harmlessness-from-ai-feedback)
- May 2023: Series C: raised $450M, led by Spark Capital [1](https://www.anthropic.com/news/anthropic-series-c)
- Sep 2023: Long-Term Benefit Trust [1](https://www.anthropic.com/news/the-long-term-benefit-trust)
- Nov 2023: Claude 2.1 released (major release) [1](https://www.anthropic.com/news/claude-2-1) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Dec 2023: Menlo-led round: in talks to raise $750M at a $18.4B valuation, led by Menlo Ventures (in talks, reported) [1](https://www.reuters.com/markets/deals/anthropic-raise-750-mln-menlo-ventures-led-funding-round-information-2023-12-21/)
- Mar 2024: Claude 3 model family released (major release) [1](https://www.anthropic.com/news/claude-3-family) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Mar 2024: Claude 3 Haiku released [1](https://www.anthropic.com/news/claude-3-haiku) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Jun 2024: Claude 3.5 Sonnet released (major release) [1](https://www.anthropic.com/news/claude-3-5-sonnet) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Oct 2024: Upgraded Claude 3.5 Sonnet and new Claude 3.5 Haiku announced [1](https://www.anthropic.com/news/3-5-models-and-computer-use) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Feb 2025: Claude 3.7 Sonnet and Claude Code released (major release) [1](https://www.anthropic.com/news/claude-3-7-sonnet) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Mar 2025: Series E: raised $3.5B at a $61.5B valuation, led by Lightspeed Venture Partners [1](https://www.anthropic.com/news/anthropic-raises-series-e-at-usd61-5b-post-money-valuation)
- May 2025: Claude Opus 4 and Claude Sonnet 4 released (major release) [1](https://www.anthropic.com/news/claude-4) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Aug 2025: Claude Opus 4.1 released [1](https://www.anthropic.com/news/claude-opus-4-1) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Sep 2025: Series F: raised $13B at a $183B valuation, led by ICONIQ [1](https://www.anthropic.com/news/anthropic-raises-series-f-at-usd183b-post-money-valuation)
- Sep 2025: Claude Sonnet 4.5 released [1](https://www.anthropic.com/news/claude-sonnet-4-5) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Oct 2025: Claude Haiku 4.5 released [1](https://www.anthropic.com/news/claude-haiku-4-5) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Oct 2025: Office: Tokyo [1](https://www.anthropic.com/news/opening-our-tokyo-office)
- Nov 2025: Claude Opus 4.5 released (major release) [1](https://www.anthropic.com/news/claude-opus-4-5) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Feb 2026: Series G: raised $30B at a $380B valuation, led by GIC and Coatue [1](https://www.anthropic.com/news/anthropic-raises-30-billion-series-g-funding-380-billion-post-money-valuation)
- Feb 2026: Claude Opus 4.6 released [1](https://www.anthropic.com/news/claude-opus-4-6) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Feb 2026: Office: Bengaluru [1](https://www.anthropic.com/news/bengaluru-office-partnerships-across-india)
- Feb 2026: Claude Sonnet 4.6 released [1](https://www.anthropic.com/news/claude-sonnet-4-6) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Mar 2026: Office: Sydney (announced) [1](https://www.anthropic.com/news/sydney-fourth-office-asia-pacific)
- Apr 2026: Claude Opus 4.7 released [1](https://www.anthropic.com/news/claude-opus-4-7) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- May 2026: Series H: raised $65B at a $965B valuation, led by Altimeter, Dragoneer, Greenoaks and Sequoia [1](https://www.anthropic.com/news/series-h)
- May 2026: Claude Opus 4.8 released [1](https://www.anthropic.com/news/claude-opus-4-8) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Jun 2026: IPO filing (reported) [1](https://www.reuters.com/business/ai-giant-anthropic-confidentially-files-us-ipo-2026-06-01/)
- Jun 2026: Office: Seoul [1](https://www.anthropic.com/news/seoul-office-partnerships-korean-ai-ecosystem)
- Jun 2026: Claude Sonnet 5 released (major release) [1](https://www.anthropic.com/news/claude-sonnet-5) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Jul 2026: Claude Opus 5 released [1](https://www.anthropic.com/news/claude-opus-5) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Sep 2026: Claude Opus 5.5 released (major release) [1](https://www.anthropic.com/claude-opus-5-5) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Sep 2026: Claude Sonnet 5.5 released [1](https://www.anthropic.com/claude-sonnet-5-5) [2](https://en.wikipedia.org/wiki/Claude_(language_model))
- Oct 2026: Claude Haiku 5.5 released [1](https://www.anthropic.com/claude-haiku-5-5) [2](https://en.wikipedia.org/wiki/Claude_(language_model))

## Things you might not know

- **It is a public benefit corporation.** Anthropic is a Delaware public benefit corporation. A separate Long-Term Benefit Trust, with no financial stake, holds a special class of stock (Class T) that lets it elect a growing share of the board. [1](https://www.anthropic.com/news/the-long-term-benefit-trust)
- **Reed Hastings is on the board.** Reed Hastings joined the board in May 2025, appointed by the Long-Term Benefit Trust. The board also includes Dario and Daniela Amodei, Yasmin Razavi, Chris Liddell and Vas Narasimhan. [1](https://www.anthropic.com/news/reed-hastings) [2](https://www.anthropic.com/company)
- **Where the name Claude may come from.** Some employees treat the name as a nod to Claude Shannon, the father of information theory, according to press reports. Anthropic has not given an official explanation. [1](https://en.wikipedia.org/wiki/Anthropic)
- **Claude can end abusive conversations.** Since August 2025, Claude Opus 4 and 4.1 can end a rare set of persistently abusive chats, an outcome of Anthropic's research program on model welfare. [1](https://www.anthropic.com/research/end-subset-conversations) [2](https://www.anthropic.com/research/exploring-model-welfare)
- **Claude Code reached a $1 billion run-rate.** Claude Code reached $1 billion in run-rate revenue in November 2025, six months after it became available to the public. Anthropic then bought Bun, the JavaScript runtime. [1](https://www.anthropic.com/news/anthropic-acquires-bun-as-claude-code-reaches-usd1b-milestone)
- **It uses a SpaceX data center.** In May 2026 Anthropic agreed to use all of the compute at SpaceX's Colossus 1 data center, more than 300 megawatts and over 220,000 Nvidia GPUs. [1](https://www.anthropic.com/news/higher-limits-spacex)
- **Claude has no ads.** In February 2026 Anthropic said Claude will carry no ads, sponsored links or paid product placement. [1](https://www.anthropic.com/news/claude-is-a-space-to-think)
- **Claude's constitution is public.** The constitution that guides Claude's training was rewritten in January 2026 and released in full under a CC0 public-domain license. [1](https://www.anthropic.com/news/claude-new-constitution)
- **Looking inside the model.** Anthropic's 2023 Towards Monosemanticity paper split a layer of 512 neurons into more than 4,000 more interpretable features. [1](https://www.anthropic.com/research/towards-monosemanticity-decomposing-language-models-with-dictionary-learning)

## Research papers the founders wrote before Anthropic

- Concrete Problems in AI Safety (2016): 3 of 6 authors later joined Anthropic. https://arxiv.org/abs/1606.06565
- Deep Reinforcement Learning from Human Preferences (2017): 3 of 6 authors later joined Anthropic. https://arxiv.org/abs/1706.03741
- The Building Blocks of Interpretability (2018): 2 of 7 authors later joined Anthropic. https://distill.pub/2018/building-blocks/
- Fine-Tuning Language Models from Human Preferences (2019): 4 of 8 authors later joined Anthropic. https://arxiv.org/abs/1909.08593
- Language Models are Unsupervised Multitask Learners (2019): 2 of 6 authors later joined Anthropic. https://cdn.openai.com/better-language-models/language_models_are_unsupervised_multitask_learners.pdf
- Language Models are Few-Shot Learners (2020): 10 of 31 authors later joined Anthropic. https://arxiv.org/abs/2005.14165
- Learning to Summarize from Human Feedback (2020): 3 of 9 authors later joined Anthropic. https://arxiv.org/abs/2009.01325
- Scaling Laws for Autoregressive Generative Modeling (2020): 8 of 19 authors later joined Anthropic. https://arxiv.org/abs/2010.14701
- Scaling Laws for Neural Language Models (2020): 6 of 10 authors later joined Anthropic. https://arxiv.org/abs/2001.08361
- Zoom In: An Introduction to Circuits (2020): 2 of 6 authors later joined Anthropic. https://distill.pub/2020/circuits/zoom-in/
Every author: https://www.autumn.ai/resources/anthropic/papers

More: the people (https://www.autumn.ai/resources/anthropic/people), where they came from (https://www.autumn.ai/resources/anthropic/talent), the research papers (https://www.autumn.ai/resources/anthropic/papers) and the money (https://www.autumn.ai/resources/anthropic/ipo).