Anthropic — research brief

Researched: 2026-09-20 (CT)
Primary focus: Claude freemium → Pro/Max consumer ladder → Team/Enterprise seats → Claude API token platform + cloud partners.

Labels: FACT = Anthropic primary page; SECONDARY = press; INFERENCE = synthesis from facts.

Company & mission

FACT — Anthropic is an AI safety and research company. Mission framing: build reliable, interpretable, and steerable AI systems. Source: Company.
FACT — Anthropic is a Public Benefit Corporation whose purpose is the responsible development and maintenance of advanced AI for the long-term benefit of humanity. Board elected by stockholders and the Long-Term Benefit Trust (LTBT). Source: Company — Governance.
FACT — Board of Directors (listed): Dario Amodei, Daniela Amodei, Yasmin Razavi, Reed Hastings, Chris Liddell, Vas Narasimhan. LTBT Trustees: Neil Buddy Shah, Richard Fontaine, Dr. Ben Bernanke. Source: Company.
FACT — Values emphasize safety as a science, “helpful, honest, and harmless,” and igniting a “race to the top” on safety. Source: Company.

Products (what they sell)

FACT — Consumer/product surface: Claude on web, desktop, iOS, Android — Free, Pro, Max; business: Team, Enterprise (self-serve and sales-assisted); developer: Claude API / Claude Platform with models Fable, Opus, Sonnet, Haiku; Claude Code, Claude Design/Slides/Docs, connectors, Managed Agents. Sources: Pricing, Claude, Models overview.
FACT — Current API flagship lineup (standard rates): Fable 5.1 $10 / $50 per MTok (input/output); Opus 5 $5 / $25; Sonnet 5 $2 / $10; Haiku 4.5 $1 / $5. Prompt caching and batch discounts documented. Sources: Pricing, docs models, docs pricing.
FACT — Models also available via Amazon Bedrock, Google Cloud, and Microsoft Foundry (platform IDs listed on models page). Source: Models overview.
FACT — Platform add-ons called out on pricing: Managed Agents $0.08 per session-hour active runtime (tokens still at model rates); code execution $0.05/hour/container after 50 free hours/day/org; US-only inference at 1.1× input/output. Source: Pricing.

Claude consumer pricing ladder

FACT — Free: $0 — chat on web/desktop/mobile; web search, files, code, memory, Artifacts, connectors. Source: Pricing.
FACT — Pro: $17/mo with annual subscription ($200 billed up front) or $20 if billed monthly. Includes Claude Code, more usage, Projects, Claude Design/Slides/Docs, Claude Science, Claude in Chrome and Microsoft 365. Source: Pricing.
FACT — Max: from $100/mo — choose 5× or 20× more usage than Pro; higher output limits; early access; priority at high traffic. Billed monthly. Source: Pricing.
FACT — Usage limits reset on a rolling five-hour session window; paid plans also have weekly limits; activity across web/desktop/mobile/Claude Code shares one pool. Pro ≈ ≥5× Free per 5-hour session; Max is 5× or 20× Pro. Source: Pricing FAQ.

Team & Enterprise

FACT — Team (2–150 seats): Standard $20/seat/mo annual or $25 monthly; Premium $100/seat/mo annual or $125 monthly. Includes SSO, central billing, Claude Code, enterprise search, connectors, no model training on content by default. Sources: Help — Team plan, Pricing / Team.
FACT — On Team, Standard seats give more usage than Pro; Premium seats give 5× more than Standard. Extra usage at standard API rates with spend controls. Source: Pricing FAQ.
FACT — Enterprise: $20 per seat per month plus usage billed at API rates (annual). Includes Team features plus SCIM, audit logs, custom data retention, RBAC, HIPAA-ready offering, Compliance API; self-serve or sales-assisted (AWS Marketplace, invoicing/net terms, CS at spend thresholds). Source: Pricing FAQ.

Partners & GTM

FACT — Cloud distribution: Amazon Bedrock, Google Cloud, Microsoft Foundry model IDs published. Source: Models overview.
FACT — Product embeds: Claude in Chrome, Claude for Microsoft 365 / Outlook, connectors / MCP, Claude Tag in Slack (Team/Enterprise beta per product page). Source: Claude, Pricing.
INFERENCE — Dual GTM like peers: Claude freemium/subscription grows brand and habit; Team/Enterprise seats + API tokens monetize orgs and builders; cloud marketplaces extend enterprise channel.

Metrics we are NOT inventing

No official Anthropic ARR, valuation, headcount, or Claude MAU/WAU cited on the primary pages fetched for this brief on 2026-09-20. Do not invent financing rounds, user counts, or revenue.
Max 20× exact dollar amount is “from $100” for the Max family on the public pricing cards; treat higher Max SKUs as usage multiples of Pro unless Anthropic lists a separate sticker price.

Teaching arc (for canvas)

Claude freemium → Pro ($17–20) / Max (from $100) → Team seats ($20/$100 annual) → Enterprise ($20 + API usage) → Claude API tokens (Fable/Opus/Sonnet/Haiku) via first-party + Bedrock/GCP/Foundry.

Free widens the funnel; Pro/Max convert power individuals; Team/Enterprise seats + API usage convert orgs; safety brand and PBC/LTBT governance differentiate trust for regulated buyers.

# Anthropic — research brief Researched: 2026-09-20 (CT) Primary focus: Claude freemium → Pro/Max consumer ladder → Team/Enterprise seats → Claude API token platform + cloud partners. Labels: **FACT** = Anthropic primary page; **SECONDARY** = press; **INFERENCE** = synthesis from facts. --- ## Company & mission - **FACT** — Anthropic is an AI safety and research company. Mission framing: build reliable, interpretable, and steerable AI systems. Source: [Company](https://www.anthropic.com/company). - **FACT** — Anthropic is a **Public Benefit Corporation** whose purpose is the responsible development and maintenance of advanced AI for the long-term benefit of humanity. Board elected by stockholders and the **Long-Term Benefit Trust (LTBT)**. Source: [Company — Governance](https://www.anthropic.com/company). - **FACT** — Board of Directors (listed): Dario Amodei, Daniela Amodei, Yasmin Razavi, Reed Hastings, Chris Liddell, Vas Narasimhan. LTBT Trustees: Neil Buddy Shah, Richard Fontaine, Dr. Ben Bernanke. Source: [Company](https://www.anthropic.com/company). - **FACT** — Values emphasize safety as a science, “helpful, honest, and harmless,” and igniting a “race to the top” on safety. Source: [Company](https://www.anthropic.com/company). ## Products (what they sell) - **FACT** — Consumer/product surface: **Claude** on web, desktop, iOS, Android — Free, Pro, Max; business: **Team**, **Enterprise** (self-serve and sales-assisted); developer: **Claude API** / Claude Platform with models Fable, Opus, Sonnet, Haiku; **Claude Code**, Claude Design/Slides/Docs, connectors, Managed Agents. Sources: [Pricing](https://www.anthropic.com/pricing), [Claude](https://www.anthropic.com/claude), [Models overview](https://docs.anthropic.com/en/docs/about-claude/models). - **FACT** — Current API flagship lineup (standard rates): **Fable 5.1** $10 / $50 per MTok (input/output); **Opus 5** $5 / $25; **Sonnet 5** $2 / $10; **Haiku 4.5** $1 / $5. Prompt caching and batch discounts documented. Sources: [Pricing](https://www.anthropic.com/pricing), [docs models](https://docs.anthropic.com/en/docs/about-claude/models), [docs pricing](https://docs.anthropic.com/en/docs/about-claude/pricing). - **FACT** — Models also available via **Amazon Bedrock**, **Google Cloud**, and **Microsoft Foundry** (platform IDs listed on models page). Source: [Models overview](https://docs.anthropic.com/en/docs/about-claude/models). - **FACT** — Platform add-ons called out on pricing: Managed Agents **$0.08 per session-hour** active runtime (tokens still at model rates); code execution **$0.05/hour/container** after 50 free hours/day/org; US-only inference at **1.1×** input/output. Source: [Pricing](https://www.anthropic.com/pricing). ## Claude consumer pricing ladder - **FACT** — **Free**: $0 — chat on web/desktop/mobile; web search, files, code, memory, Artifacts, connectors. Source: [Pricing](https://www.anthropic.com/pricing). - **FACT** — **Pro**: **$17/mo** with annual subscription ($200 billed up front) or **$20** if billed monthly. Includes Claude Code, more usage, Projects, Claude Design/Slides/Docs, Claude Science, Claude in Chrome and Microsoft 365. Source: [Pricing](https://www.anthropic.com/pricing). - **FACT** — **Max**: from **$100/mo** — choose **5× or 20×** more usage than Pro; higher output limits; early access; priority at high traffic. Billed monthly. Source: [Pricing](https://www.anthropic.com/pricing). - **FACT** — Usage limits reset on a rolling **five-hour** session window; paid plans also have weekly limits; activity across web/desktop/mobile/Claude Code shares one pool. Pro ≈ **≥5×** Free per 5-hour session; Max is 5× or 20× Pro. Source: [Pricing FAQ](https://www.anthropic.com/pricing). ## Team & Enterprise - **FACT** — **Team** (2–150 seats): Standard **$20/seat/mo annual** or **$25 monthly**; Premium **$100/seat/mo annual** or **$125 monthly**. Includes SSO, central billing, Claude Code, enterprise search, connectors, no model training on content by default. Sources: [Help — Team plan](https://support.anthropic.com/en/articles/9266767-what-is-the-claude-team-plan), [Pricing / Team](https://www.anthropic.com/pricing). - **FACT** — On Team, Standard seats give more usage than Pro; Premium seats give **5×** more than Standard. Extra usage at standard API rates with spend controls. Source: [Pricing FAQ](https://www.anthropic.com/pricing). - **FACT** — **Enterprise**: **$20 per seat per month plus usage billed at API rates** (annual). Includes Team features plus SCIM, audit logs, custom data retention, RBAC, HIPAA-ready offering, Compliance API; self-serve or sales-assisted (AWS Marketplace, invoicing/net terms, CS at spend thresholds). Source: [Pricing FAQ](https://www.anthropic.com/pricing). ## Partners & GTM - **FACT** — Cloud distribution: Amazon Bedrock, Google Cloud, Microsoft Foundry model IDs published. Source: [Models overview](https://docs.anthropic.com/en/docs/about-claude/models). - **FACT** — Product embeds: Claude in Chrome, Claude for Microsoft 365 / Outlook, connectors / MCP, Claude Tag in Slack (Team/Enterprise beta per product page). Source: [Claude](https://www.anthropic.com/claude), [Pricing](https://www.anthropic.com/pricing). - **INFERENCE** — Dual GTM like peers: Claude freemium/subscription grows brand and habit; Team/Enterprise seats + API tokens monetize orgs and builders; cloud marketplaces extend enterprise channel. ## Metrics we are NOT inventing - No official Anthropic ARR, valuation, headcount, or Claude MAU/WAU cited on the primary pages fetched for this brief on 2026-09-20. Do **not** invent financing rounds, user counts, or revenue. - Max 20× exact dollar amount is “from $100” for the Max family on the public pricing cards; treat higher Max SKUs as usage multiples of Pro unless Anthropic lists a separate sticker price. ## Teaching arc (for canvas) **Claude freemium → Pro ($17–20) / Max (from $100) → Team seats ($20/$100 annual) → Enterprise ($20 + API usage) → Claude API tokens (Fable/Opus/Sonnet/Haiku) via first-party + Bedrock/GCP/Foundry.** Free widens the funnel; Pro/Max convert power individuals; Team/Enterprise seats + API usage convert orgs; safety brand and PBC/LTBT governance differentiate trust for regulated buyers.