led by Dario Amodei

Anthropic profile, products and governance

Builders of Claude, the AI model family spanning Fable, Opus, Sonnet and Haiku.

Hokai's read on Anthropic

Anthropic suits regulated buyers in finance, healthcare and government that want a vendor with a published safety framework and independent attestations such as SOC 2 and ISO 42001. The trade-off is reach: Claude.ai trails ChatGPT among consumers, and its record Series H valuation rests on revenue that analysts still debate.

Anthropic products

Models from Anthropic

ModelLabContextIn / out per 1MReads and writes
Claude Fable 5.2Anthropic's November 2026 Fable model: the Mythos 5.2 weights behind cyber and biology safeguards, built for multi-hour agentic coding and research.Anthropic1M$8 / $40text, image → text
Claude Mythos 5.2Anthropic's November 2026 Mythos model: Fable 5.2's weights with relaxed cyber and biology safeguards for verified teams, with a 1M-token context.Anthropic1M$8 / $40text, image → text
Claude Haiku 5.5Anthropic's smallest and fastest 5.5 model, released October 7, 2026, with a 1M context window, built for high-volume work.Anthropic1M$0.10 / $0.50text, image → text
Claude Sonnet 5.5Anthropic's fastest Sonnet-class model, launched in September 2026, with a 1M context and output more than 30% faster than Sonnet 5.Anthropic1M$2 / $10text, image, pdf → text
Claude Opus 5.5Anthropic's September 2026 flagship LLM with a 1 million token context window, built for long-running agentic coding and computer use.Anthropic1M$4 / $20text, image, pdf → text
Claude Fable 5.1Superseded by Claude Fable 5.2 in November 2026 but still sold on the Claude API: Anthropic's second Mythos-class model, released Sept 1, 2026, for long-horizon coding and research.Anthropic1M$10 / $50text, image, pdf → text
Claude Mythos 5.1Anthropic's Mythos-class model, identical to Fable 5.1 but with safeguards relaxed for verified cyber and life-sciences users, scoring 1,853 on GDPval-AA v2.Anthropic1M$10 / $50text, image → text
Claude Opus 5Anthropic's July 2026 Opus with a 1M token context and xhigh reasoning effort, still active but superseded by Claude Opus 5.5 on September 22, 2026.Anthropic1M$5 / $25text, image, pdf → text
Claude Sonnet 5Claude Sonnet 5, released by Anthropic on June 30, 2026, hits 82.1% SWE-bench Verified with a 1M-token context window.Anthropic1M$3 / $15text, image, pdf → text
Claude Fable 5Superseded by Fable 5.1 and Fable 5.2 but still sold on the Claude API: Anthropic's first generally available Mythos-class model, released June 2026, with a 1M-token context.Anthropic1M$10 / $50text, image, pdf → text
Claude Mythos 5Anthropic's verification-only Mythos-class model: Claude Fable 5's base model with cybersecurity safeguards lifted for vetted defenders. A successor took over as the current Mythos model on 1 September 2026, though Anthropic still lists this one as active.Anthropic1M$10 / $50text, image, pdf → text
Claude Opus 4.8Claude Opus 4.8 (May 2026) is Anthropic's former flagship, now a legacy release with a 1M-token context window. Its newer successor is the recommended choice for new builds.Anthropic1M$5 / $25text, image, pdf → text
Claude Opus 4.7Claude Opus 4.7 by Anthropic (April 2026): 1M-token context, 87.6% SWE-bench Verified, high-res vision, and adaptive thinking at $5/$25 per 1M tokens.Anthropic1M$5 / $25text, image, pdf → text
Claude Sonnet 4.6Claude Sonnet 4.6 by Anthropic (Feb 2026) scores 79.6% on SWE-bench Verified with a 1M-token context window, and is now one generation behind Sonnet 5.5.Anthropic1M$3 / $15text, image, pdf → text
Claude Opus 4.6Claude Opus 4.6, released February 5, 2026, has a 1M-token context window and 128k max output, priced at $5/$25 per 1M tokens and listed Active by Anthropic.Anthropic1M$5 / $25text, image, pdf → text
Claude 1.0Claude 1.0 was Anthropic's first widely offered model: text-only, trained with Constitutional AI, and since retired. This page records what it was and what replaced it.Anthropic9K--text → text

Tools from Anthropic

ToolPrice fromFree planRuns onRating
Claude CodeAnthropic's agent for the terminal, IDEs, desktop, web and mobile that edits files, runs tests and opens pull requests. There is no free plan.$20/moNoWeb · iOS · Android · Mac · Windows4.6
ClaudeClaude is Anthropic's AI assistant, with a free tier, Pro, Max and Ultra personal plans, and Team and Enterprise seats. Opus access starts at Pro.$20/moYesWeb · iOS · Android · Mac · Windows4.3

Agents from Anthropic

AgentWhere it runsPriceMaturityRating
Claude DuetAnthropic's proactive agent that briefs you daily and asks before it actsCloud$20/moAlpha3.6

Services from Anthropic

ServiceRegionsPricingComplianceChecked
AnthropicClaude models via API, Claude.ai plans, Bedrock and Vertex AI.US · Europe · Asia PacificFree tier · Pay as you go · SubscriptionSOC 2 · ISO 27001 · ISO 42001 +19 Oct

Key products, as the company lists them

ProductWhat it is
Claude.aiConsumer chat product
Claude Developer PlatformDeveloper API
Claude for EnterpriseB2B SAAS
Claude CodeCLI / agentic dev tool
Claude CoworkMulti-agent coordination
Claude ScienceResearch workbench
Claude DesignDesign product
Claude SecuritySecurity product
Claude in ChromeBrowser extension
Claude DuetProactive agent (research preview)
Claude TagSlack agent

Anthropic facts

Company type
Research lab
Founded
2021
Headquarters
San Francisco, California, USA
Led by
Dario Amodei
Valuation
$965B post-money (Series H, May 28, 2026)
Trust centre
Published
Privacy policy
Published
Terms of service
Published
Team size
3,400-5,000 (private company, no official figure)
Funding
$132B total raised across 18 rounds; latest $65B Series H (May 28, 2026) led by Altimeter Capital, Dragoneer, Greenoaks, Sequoia Capital

Compliance certifications

  • SOC 2 Type I
  • SOC 2 Type II
  • ISO 27001:2022
  • ISO/IEC 42001:2023
  • HIPAA-eligible
  • GDPR

Data regions

  • United States (AWS, Google Cloud, Microsoft Azure)
  • European Union (GDPR data residency options)

Anthropic releases

Every model Hokai holds for this company that carries a release date, newest first.

Show all 16 releases

Companies like Anthropic

CompanyTypeBased inFoundedOn Hokai
Anthropic This pageResearch labSan Francisco, USA202120
OpenAIPick Anthropic for its public safety framework and enterprise compliance certifications; pick OpenAI for broader consumer reach and multimodal features like image and voice.Research labSan Francisco, USA201527
GooglePick Anthropic for a safety-first research focus; pick Google for native multimodal capability and default distribution through Android and Search.Big techMountain View, USA199827
Mistral AIPick Anthropic for a closed, safety-audited model; pick Mistral for open-weight models and European data sovereignty.Research labParis, France20239

Also compared

  • CoherePick Anthropic for general-purpose reasoning across consumer and developer products; pick Cohere for RAG-optimized private enterprise deployment.

Questions about Anthropic

How much does Claude cost?

Claude.ai has a Free plan, Pro at $17 a month on annual billing ($20 billed monthly), Max at $100 a month and Ultra at $200 a month, read from claude.com/pricing on 7 November 2026. Team standard seats are $20 a month on annual billing, and the Claude API bills per token by model. Check the pricing page before budgeting.

Is there a free way to try Claude?

Yes. The Free plan costs nothing and gives limited usage every 5 hours and every week, with web search, file creation, code execution and memory across conversations. Developers can test the Claude API through the Claude Developer Platform before committing to metered usage.

Which companies compete with Anthropic?

OpenAI is the closest rival on consumer chat and some benchmarks, and Google DeepMind competes on multimodal features and Android and Search distribution. Mistral AI and Cohere target open-weight models and private RAG-style enterprise deployment. Anthropic differs by selling closed models with a published safety framework.

Is Anthropic worth more than OpenAI, and is it public yet?

Anthropic reported a $965 billion post-money valuation in its May 28, 2026 Series H, which put it above OpenAI for the first time. It is still private: it filed an S-1 on October 26, 2026 and plans a Nasdaq listing under the symbol ANT, with the date and price range not yet set. OpenAI still has the larger consumer audience through ChatGPT.

Where do I start if I want to build on Claude?

Sign up at claude.ai to chat right away, or open the Claude Developer Platform at platform.claude.com for API keys and docs. Teams already on AWS, Google Cloud or Azure can reach Claude through Bedrock, Vertex AI or Azure without opening a separate Anthropic account.

About Anthropic

Anthropic is a San Francisco Public Benefit Corporation founded in 2021 by seven former OpenAI researchers. It builds the Claude model family, which as of 7 November 2026 spans Fable 5.1, Opus 5.5, Sonnet 5.5 and Haiku 5.5, and it filed publicly for a Nasdaq IPO in October 2026.

In their words

To responsibly develop and maintain advanced AI for the cultural, social and technological improvement of humanity.

Anthropic was incorporated as a Public Benefit Corporation in San Francisco in January 2021 by seven ex-OpenAI researchers who split over safety: Dario Amodei (CEO), Daniela Amodei (President), Tom Brown, Jack Clark, Jared Kaplan, Chris Olah, and Sam McCandlish. HQ is 548 Market Street, San Francisco, with offices in Seattle, New York, DC, London, Paris, Munich, Seoul (since June 2026), and Singapore (opened October 14, 2026, with a memorandum of understanding signed with the Singapore government).

Read the full background

The flagship product is Claude.ai, alongside the Claude API and Claude for Enterprise. Anthropic's own newsroom shows the Claude family moving fast through autumn 2026: Opus 5.5 on September 22, Sonnet 5.5 on September 28 (described by Anthropic as 30 percent faster and up to 30 percent cheaper for most work than Sonnet 5), and Haiku 5.5 on October 7, billed as its fastest and cheapest small model yet. Claude Fable 5.1 sits above them, and Mythos 5 remains in preview. Anthropic says Opus 5.5 performs at the level of Fable 5.1 on most work at 40 percent lower running cost than Opus 5. The products around the models include Claude Code, Claude Cowork, Claude Design, Claude Science, Claude Security, Claude in Chrome and Claude Tag, which lets teams call Claude from any Slack thread. For help choosing between the tiers, see the guide on which Claude model to use.

Anthropic has raised roughly 132 billion dollars across 18 rounds: 61.5 billion in March 2025, 183 billion in September 2025, a 30 billion dollar Series G in February 2026 at 380 billion, and a 65 billion dollar Series H on May 28, 2026, at 965 billion, ahead of OpenAI for the first time. The Series H was led by Altimeter, Dragoneer, Greenoaks, and Sequoia. Amazon (8 billion plus, Project Rainier) and Google (equity plus a TPU deal) are the largest backers. Anthropic confidentially submitted a draft S-1 on June 1, 2026. On October 26, 2026 it filed a registration statement with the SEC for an initial public offering and said it intends to list Class A shares on the Nasdaq Global Select Market under the symbol ANT. The share count and price range are not set and the registration is not yet effective, so Anthropic is still private and has announced no listing date.

Revenue is mostly API usage billed per token, split between enterprise contracts, Claude.ai subscriptions (Free, Pro, Max, Team), and reseller access through Bedrock, Vertex AI, and Azure, plus Claude Code. Run-rate revenue climbed from roughly 9 billion dollars at 2025 year-end to over 47 billion by May 2026, a figure analysts are watching because gross run-rate and net recognized revenue can differ.

Anthropic does not publish an official headcount; estimates range from about 3,400 at end-2025 to over 5,000 by mid-2026, up from around 500 in 2023. Mike Krieger, Instagram co-founder and CPO since 2024, moved in 2026 to co-lead Anthropic Labs. Chris Olah leads interpretability; Jan Leike leads alignment.

The stated public benefit purpose is to responsibly develop advanced AI for humanity's cultural, social and technological improvement. Constitutional AI, published December 2022, trains models against a written set of principles scored by a critic model rather than relying only on human feedback. On June 10, 2026, Anthropic published its Advanced AI Framework, extending its earlier Responsible Scaling Policy into a formal commitment tied to capability thresholds.

October 2026 brought a run of commitments in the newsroom: a 100 million dollar Claude Frontier Academy to train 10,000 engineers by the end of 2027 (October 2), an expanded Cyber Verification Program for qualifying security professionals (October 6), and a 150 million dollar, three-year pledge of Claude access for more than 15 agencies in the US Genesis Mission science initiative (October 8). On October 18 it announced the Anthropic Public Benefit Fund, a wholly owned LLC to which, subject to stockholder approval, 5 percent of Anthropic's shares would be committed, with a further 2 percent added at the board's discretion for every 50 percent rise in company value above a set baseline. Physicist Christopher Stubbs joined the Long-Term Benefit Trust on October 21.

The closest rival is OpenAI, whose ChatGPT still has the larger consumer audience. Google DeepMind competes on multimodal capability and default distribution via Android and Search. Where the alternative wins: a buyer who needs the widest consumer reach, native image and voice generation, or open-weight models for self-hosting (see Mistral AI) will find Anthropic's closed, safety-audited approach a poorer fit, and Claude.ai trails ChatGPT's user base. Anthropic's edge is enterprise trust: its PBC status, safety framework, and compliance certifications make it a default pick for regulated buyers in finance, healthcare, and government. Its systemic-risk-level models put it under the EU AI Act's stricter general-purpose-AI rules, and in the US it worked with export-control authorities after a directive briefly forced it to disable Fable 5 and Mythos 5 for every customer on June 12, 2026, before access was restored. Infrastructure spans AWS's Project Rainier campus in Indiana, expanded Google Cloud TPU capacity, a new Azure and Nvidia deal, and a 50 billion dollar Fluidstack data center build in Texas and New York. Anthropic also sells model access as a cloud product; see its Anthropic service listing for plans and regions.

Sources and guides

More AI Companies on Hokai

  • NdeaAI research lab betting AGI comes from program synthesis, not more scale.
  • Nex AGIOpen-weight agent models and the NexAU framework, from Shanghai
  • xAIElon Musk's AI lab building Grok, now SpaceX's AI arm with Cursor and Grok Bot.
  • Google DeepMindAlphabet's AI research lab behind Gemini, Gemma and AlphaFold.
  • Z.aiChina's first publicly listed AI lab, open-sourcing the GLM model family.
  • Hanno LabsApplied AI for the work that still requires expert judgment