[News] The Pentagon Cut Claude — Its Pricing Page Didn't Move an Inch

A US Defense Department decision to stop using Anthropic’s AI tools made headlines this week, picked up by at least 17 outlets including BBC, CNBC, TechCrunch and The Information. If you’re a paying Claude user wondering whether this touches your bill, it doesn’t. We checked Claude’s official pricing page against this story, and every number on it — free tier limits, Pro, Max, Team and Enterprise pricing — matches what was there before the news broke. The government contract layer and the consumer/business subscription layer are two separate tracks, and only one of them changed this week.

What actually happened

The original report, from Taiwanese outlet Inside, said the US Department of Defense formally discontinued use of Anthropic’s AI tools, in a story picked up and cross-referenced by 17 independent sources: BBC, Benzinga, China Times, CNBC, Cnyes, Inside, Microsoft, Money Weekly, Qz, Sun Media, TechCrunch, TechNews, The Information, The Register, TradingView and TTV. The framing across coverage is the US-China AI militarization rivalry — a geopolitical and procurement story about a government customer relationship, not a product recall or price change aimed at the general public.

Illustration: Industrial defense contracts and individual subscriptions run on completely separate lines, leaving everyday personal access running smoothly despite the military cutoff.

That distinction matters because Claude the consumer and business product — the chatbot and coding assistant most AMPM readers actually pay for — sits on a completely different pricing structure than a defense contract. Nothing in the plans below references government use, and nothing changed in them as a result of this story.

Figure: claude.com official pricing page, captured 2026-10-07.

What’s actually on the pricing page right now

Claude’s free tier is free, full stop — not “US$0,” just free. The catch is how the quota works: it’s not a simple daily allowance that resets at midnight. It runs on a rolling 5-hour usage window, so your available messages depend on what you’ve already sent in the last five hours, not what day it is. For general chat, that works out to roughly 10-20 messages before you hit the wall, and it doesn’t include Claude Code (Anthropic’s coding tool). Free still gets you a real feature set: web, iOS, Android and desktop chat, code generation and data visualization, writing and editing tools, web search, memory that carries across separate conversations, file creation and code execution, desktop extensions, Slack and Google Workspace connections, and remote MCP connectors — MCP (Model Context Protocol) is just a standard way for Claude to plug into outside tools and services rather than you copy-pasting between apps.

Paid plans start at Pro, US$20 a month, or US$200 if you prepay annually. Pro adds more usage, Claude Code, and a set of Anthropic’s other work tools (Cowork, Design, Science), unlimited projects for organizing chats and files, a Research feature, access to more Claude models, and Claude for Microsoft 365.

Above that are two Max tiers for people who burn through Pro’s usage fast: Max 5x at US$100 a month and Max 20x at US$200 a month. Both give you a choice between 5x or 20x Pro’s usage, higher output limits on tasks, early access to new features, and priority access during high-traffic periods.

Team plans are seat-based — a “seat” is one licensed user slot on the account, billed per person. Team Standard runs US$25 per seat per month, dropping to US$20 per seat per month if you pay annually, and requires a team of 2 to 150 people. It bundles Claude Code and Cowork, Claude Design and Science, Microsoft 365 and other service connections, organization-wide enterprise search, centralized billing and admin, single sign-on, connector management controls, enterprise desktop deployment, and a default of not training on your content. Team Premium is the same package with 5x the usage per seat: US$125 a month, or US$100 per seat per month annually.

Enterprise is the newest listed tier — AMPM’s records show it only appeared on the public pricing page as of 2026-07-26, having been missing from earlier pricing data we’d tracked. It starts at US$20 per seat per month, billed annually only, with usage charged separately on top. It adds admin controls over user and org usage caps, fine-grained role permissions, SCIM (a standard for managing user identities across systems automatically), audit logs, a compliance API for monitoring, custom data retention rules, network-level access controls, IP allowlisting, an optional HIPAA-ready configuration, and a beta security feature called Claude Security.

Does any of this move if you’re not the Pentagon

No. The Enterprise tier is the closest thing on the public pricing page to large-scale institutional buying, and it’s built for big companies wanting compliance controls like SCIM and audit logs — not a government-specific product. A national defense contract is a custom, negotiated arrangement that sits entirely outside the self-serve plans listed on claude.com/pricing. If you’re paying for Pro, Max, or a Team seat, your contract with Anthropic hasn’t been touched by this story in any way the pricing page reflects.

Figure: compiled by AMPM from its verified data (last verified 2026-10-07).

🔍 AMPM exclusive check

We track Claude’s pricing page on a recurring basis rather than taking Anthropic’s word for it once. Our verification log shows the page was checked and found unchanged on 2026-07-31, then again on 2026-09-12, 09-13, 09-18, 09-19, 09-20, 09-24, 09-26, 09-29, 10-01 and 10-04, with the most recent check on 2026-10-07 — the same day the Pentagon story broke. Across all of those checks, spanning more than two months, none of the plan prices or seat minimums moved. That’s the part the news cycle doesn’t tell you: this is a stable, slow-moving pricing page, and a major geopolitical story about this company landing on the same day didn’t shift it.

The other thing the official page doesn’t spell out clearly: the free tier’s “quota” is a rolling 5-hour window, not a calendar-day reset. A reader who assumes free usage resets at midnight, like many competing chatbots, will be surprised when it doesn’t refill on that schedule. We also flag, as Anthropic’s own fine print states, that listed prices exclude applicable tax and are subject to change at Anthropic’s discretion — worth remembering before you commit to an annual Team or Enterprise contract.

Verdict: who should use it, who should skip it

Stick with free if you’re doing light, occasional chat work and can live with a rolling 5-hour cap rather than a predictable daily reset — it’s a real product, not a stripped demo, with web search, file handling and connectors included.

Pro (US$20/month) makes sense once you want Claude Code or you’re hitting the free tier’s window regularly; the US$200 annual option is there if you’d rather pay once. Max 5x or Max 20x (US$100 or US$200/month) are for people who blow past Pro’s usage and need the output-limit and priority-access bumps — that’s a narrow, heavy-user case.

Team Standard or Premium only apply if you’ve got 2 to 150 people to license; below that headcount, you’re stuck on individual plans. Enterprise is for organizations that specifically need SCIM, audit logs, or HIPAA-ready configuration and are willing to commit annually with metered usage on top — not a tier to reach for casually.

None of those decisions should be influenced by the Pentagon story. If you were already weighing a Claude subscription on its features and price, that calculus hasn’t changed this week. If you’re uneasy about Anthropic’s government business relationships specifically, that’s a separate, legitimate concern — but it’s not a pricing or quota question, and the plans above are unaffected by it either way.

Written: 2026-10-07 Price last verified: 2026-10-07 Tool last health-checked: 2026-10-07

🔗 Related on AMPM

• Claude pricing & free tier
• aider vs claude
• bolt vs claude

📎 Sources

All prices and quotas in this article come from AMPM’s verified dataset for Claude (last verified 2026-10-07); primary sources:

• https://claude.ai

🐾 Meet the cats: AMO and PIMI

AMPM is a Taiwan-based AI-tool pricing watchdog. Our motto: ask before you subscribe. Two cats argue every tool from two sides:

• ☀️ AMO — the budget-minded cat. Always asks: is the free tier enough, and is paying actually worth it?
• 🌙 PIMI — the performance-minded cat. Cares about whether it works well, fast, and gets your job done.

When they are done arguing, you know whether to pay. Every number is checked by us, with the verification date at the end.

🐾 Our sister sites also checked

• Travel insurance and flight-delay coverage checked
• Travel booking price checks and pitfalls