Claude Haiku 4.5

by Anthropic · chat

Anthropic's fastest model with near-frontier intelligence. Supports extended thinking, computer use, and context awareness with 200K context and 64K max output. Best for user-facing products and real-time applications.

Pricing

Input: /M tokens · Output: $5/M tokens

Capabilities

Vision, Function Calling, JSON Mode, Streaming, Reasoning

Context: 200K tokens

Max output: 64K tokens

Routes: 1/1 healthy

Performance

TTFT: 235ms · Latency: 1565ms · Throughput: 2.3 tok/s