Anthropic Pricing Plans & Tiers

Anthropic's consumer pricing—Free, Pro ($17/mo), Max ($100/mo)—mirrors Perplexity and Claude.ai but embeds a critical asymmetry: Anthropic operates two separate pricing models depending on user channel. Web/mobile users pay $17/mo or $100/mo for usage-limited subscriptions, while API customers pay per-token on a metered basis. This pricing schism is sustainable because the company is optimizing for different objectives: consumer subscriptions maximize LTV through affordable recurring revenue, while API pricing maximizes enterprise revenue extraction through metered usage. Organizations building Claude-dependent products pay orders of magnitude more than individual subscribers.

Claude Pro ($17/mo) includes 'more usage' (undefined limit) but positions the tier as unlocking Model access and feature parity with Max. This ambiguity—what is 'more usage'?—exploits consumer uncertainty to force Max adoption. Teams genuinely using Claude for research and code generation discover Pro's limits within weeks and face a $100/mo step change. This pricing structure generates revenue from upgrade friction rather than from genuine feature differentiation.

The Max tier ($100/mo) includes 'Early access to advanced Claude features'—language that suggests Max is not a endpoint product but a preview tier for future capabilities. This positions the tier as a continuous commitment (to access emerging features) rather than a one-time purchase, creating psychological lock-in. Teams adopting Max for upcoming features internalize the cost and continue paying even when features arrive on lower tiers.