Claude Haiku
Anthropic's smallest, fastest, cheapest Claude tier — used for bulk classification, routing, and trivial agent steps.
Definition
Claude Haiku is the entry tier of Anthropic's model lineup, optimised for speed and unit cost. Typical uses: tag classification on incoming records, light summarisation, routing decisions inside an agent, and any high-volume call where Sonnet would be overkill. Haiku is the model most often run inside production data pipelines.
When to use
See also
- Claude Sonnet — Anthropic's mid-tier Claude model family — the default workhorse balancing speed, cost, and reasoning quality.
- Claude Opus — Anthropic's most capable Claude tier — highest reasoning quality at the highest cost and latency.
- LLM — Large Language Model — a transformer-based model trained on internet-scale text to generate and reason.