Anthropic introduces usage-based pricing for Claude Fable 5, signaling shift away from flat subscriptions

Anthropic is upending the consumer AI pricing model. Beginning July 12, users on the company's $20, $100, and $200 monthly plans must pay additional fees—$10 per million input tokens and $50 per million output tokens—to access Claude Fable 5, the consumer-grade version of its flagship Mythos 5 model. This marks the first time a frontier AI lab has applied usage-based billing to consumer access rather than flat-rate subscriptions.

Why the shift?#

The move reflects tightening computational resources. Anthropic has secured multibillion-dollar data center agreements with SpaceX, Amazon, and Google, yet demand for Fable 5 remains constrained. The company signaled it will return the model to its standard subscription plans "when sufficient capacity allows," but offered no timeline.

The pricing change also mirrors broader industry movement. Developers have long paid per-token; now consumer-facing startups like Cursor have abandoned unlimited plans for consumption-based models, citing unsustainable economics. Executives including OpenAI's Nick Turley have argued that "unlimited AI plans" no longer make business sense when agents like Claude Code consume far more compute than traditional chatbots.

Competitive positioning#

Claude is gaining traction in the consumer market. The platform reached 245 million unique monthly visitors in May, more than doubling since February, though it trails ChatGPT's 1.11 billion and Google Gemini's 662 million. By restricting free access and charging power users, Anthropic appears to be testing whether brand loyalty and quality perception justify premium pricing—essentially positioning itself as the "Apple of the AI era."

The strategy assumes users view AI models not as commodities but as status markers: choosing the "best intelligence" rather than adequate intelligence. Anthropic's explicit refusal to introduce advertising revenue may leave usage-based pricing as its primary path to profitability while the company pursues a reported initial public offering.

What's next#

Rivals like OpenAI and Google are taking different routes, favoring ad integration in free tiers. Anthropic's posture against advertising leaves it fewer levers. Whether its premium positioning can sustain higher costs amid intensifying competition—particularly from OpenAI's latest models—remains uncertain. The "golden era" of subsidized consumer AI subscriptions appears to be ending.