modelmeter.

Claude 5.5 Haiku

Anthropic

Claude 5.5 Haiku is a large language model developed by Anthropic. It is designed for understanding and generating code and complex tasks.

Token pricing and context

Input
$0.1 USD / 1M tokens
Cached input
$0.01 USD / 1M tokens
Output
$0.5 USD / 1M tokens
Context window
1000000

Benchmark results

MM1.0

Overall score: 84 · Scale 0–100 · Higher is better

  • Reasoning: 96
  • Coding: 80
  • Chat: 100
  • Long Context: 93
  • Vibe: 80
  • Accuracy: 93
  • General Knowledge: 37

Evaluated: 2026-10-17

Record exact model version, reasoning setting, temperature, tools, and evaluation setup.

Updated 2026-10-17. Read benchmark methodology.