Claude Team Plan & iOS App: Pricing, Seats and Features

A five-seat minimum and the context window as the pitch

Anthropic's May 1, 2024 announcement set Claude Team at $30 per person per month, with a minimum of five seats, so the smallest possible purchase was $150 a month. The plan bundled higher usage limits than the individual Pro plan, access to the full Claude 3 family (Opus, Sonnet and Haiku), admin tools for billing and seat management, and the 200K token context window for long documents. [1]

Of those features, the context window is the one that does not exist on the individual plan at all rather than merely being capped lower, which is why the announcement leads with it as the reason a business would pay a seat price instead of five individual subscriptions. [1]

The business plan came first, the consumer app second

The post presents Claude Team ahead of the iOS app, and closes by previewing features still in development rather than shipped that day: citations to verify AI generated claims, integrations with codebases and CRMs, and shared editing of AI generated documents with colleagues. [1]

Pairing a paid business tier with a free consumer app in one post, in that order, reads as Anthropic treating the business plan as the more consequential launch of the two, with the iOS app as a parity feature it could no longer skip rather than a headline in its own right. [1]

A year behind ChatGPT's app, and what it actually does

The iOS app shipped free to all plan types, syncing chat history across devices and adding vision capability to analyze photos taken on the phone. [1]

OpenAI had already put a ChatGPT app on iOS the previous year, rolling out first in the United States with plans to expand to more countries afterward. [2]

A native mobile app was not a capability gap since Claude's website worked in a mobile browser the whole time. What changed on May 1, 2024 was distribution: an App Store listing, a home screen icon and push notifications, the things that turn occasional use into a habit. Anthropic shipped that roughly a year after OpenAI did, on a product where the underlying model quality mattered less than being the app already installed on the phone. [1][2]