Image: Unknown · CC0 1.0 Public Domain Dedication · Openverse
Anthropic’s hosted models for text, code and tool use, billed per million tokens.
What it changes
Gives builders a frontier model through one API with prompt caching, tool calling and batch jobs, instead of serving and scaling weights themselves.
Availability
Generally available (global)
Access and cost
Get Claude API — Per-token rates are published for every model; the largest list $10 per million input tokens and $50 per million output.
Evidence it works
- Publishes per-million-token rates for every current model, with cache reads at $0.25 per million tokens on the flagship models. — source
- Lists subscription plans alongside API pricing, so buyers can compare seat pricing with metered use. — source
What it cannot do yet
- Rate limits are tiered by account spend, so a new key starts throttled until usage history exists.
- Long-context requests are billed above standard input rates, so the widest window is the most expensive path.
Anthropic · AI Models · Last verified 2026-09-17. Claims are re-checked on a 30-day clock; the date above is when this line was last confirmed.
About the Author
