← Model indexMODEL PROFILE / claude-haiku-5-5

Anthropic / current

Claude Haiku 5.5

Anthropic's fastest and cheapest small model, aimed at high-volume work.

Context1M
ModalityText · Image
AccessAPI · Cloud
ReasoningAdaptive

Where it fits.

Runs about seventy-five percent cheaper than Haiku 4.5 and is the first Haiku class model with an adjustable effort setting. Launched October 7, 2026 at $0.10 input and $0.50 output per million tokens for prompts up to 100K tokens, rising to $0.50 and $2.50 above that. Anthropic reports roughly 75% lower average run cost than Haiku 4.5 and 72.4% on the OSWorld 2.1 offline subset.

02

Core capabilities

  • 1M context window with up to 128K output
  • First Haiku class with adjustable effort setting
  • Very low cost for prompts under 100K tokens
  • Available on Claude API, Bedrock, Vertex and Azure including GovCloud
03

Best-fit workflows

  • Summaries and compaction
  • Classification and queries
  • Live support and browser use
04 / OPERATIONS

From evaluation
to production.

Deployment notes

Use as the default subagent for summaries, classification and database queries, with Sonnet 5.5 escalation. Keep the effort setting explicit and track per-task cost at both sides of the 100K threshold.

Watch before choosing

Above 100K input tokens the price rises sharply; keep prompts inside the cheaper band where possible. Cybersecurity safeguards are more restrictive than Haiku 4.5 and still block penetration testing; do not route red-team work here.

05

Family & versions

Specifications and positioning checked against provider documentation.

Open official source↗