Skip to content
HopBase Docs
Sign in
claude-haiku-5-5ChatExamples verified 2026-09-25

Claude Haiku 5.5

The latest Haiku, with 1M context and 128K max output, priced in two tiers by prompt length. Built for high-volume, latency-sensitive work such as classification, extraction, and routing.

Official price
$0.10 / $0.50/ 1M tokensInput / output
Context window
1M tokens1,000,000
Notes
  • Anthropic protocol only: the Base URL has no /v1, and /v1/chat/completions returns 404.
  • Prompts (input + cache reads + cache writes) up to 100K tokens: $0.10 / $0.50, cache hits $0.01/M. Above 100K: every rate ×5, i.e. $0.50 / $2.50, cache hits $0.05/M.
  • Adaptive thinking; control depth with output_config.effort (default medium).
Read the integration guideJSON Schema

Pricing

Official price in USD. Actual charge = official price × group multiplier.

Input
$0.10 / 1M tokens
Cached input
$0.01 / 1M tokens
Output
$0.50 / 1M tokens

Endpoints

Same family

claude-sonnet-5-5claude-sonnet-5claude-sonnet-4-6claude-opus-5-5claude-opus-5claude-opus-4-8claude-opus-4-7claude-opus-4-6claude-haiku-4-5-20251001claude-fable-5-1claude-fable-5

Prices come from the public catalog; limits come from specs exported from the gateway’s validation code. Actual charge = official price × group multiplier.