Claude models

Right-sized for any task, our models offer the best combination of speed and performance.

Sonnet 4.5

Optimal balance of intelligence, cost, and speed

Input
≤200K $3/MTok
>200K $6/MTok
Output
≤200K $15/MTok
>200K $22.5/MTok
Cache ≤200K
Write $3.75/MTok
Read $0.30/MTok
Cache >200K
Write $7.50/MTok
Read $0.60/MTok

Opus 4.5

Most intelligent for agents and coding

Input $5/MTok
Output $25/MTok
Caching
Write $6.25/MTok
Read $0.50/MTok

Haiku 4.5

Fastest, most cost-efficient model

Input $1/MTok
Output $5/MTok
Caching
Write $1.25/MTok
Read $0.10/MTok