Anthropic rolled out Claude Haiku 5.5 with lower prices than its predecessor, calling it its "cheapest, fastest, and most capable small model" capable of handling high-volume, automated enterprise workflows.
In a statement on Wednesday, the company said the model is priced around 75% below Haiku 4.5 on average and designed for tasks including summaries, database queries, classification, and subagent work. It is also intended for speed-sensitive applications such as live customer support and browser use.
Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens for prompts of up to 100,000 tokens. However, the prices rise to $0.50 and $2.50, respectively, for larger prompts, Anthropic said.
Haiku 5.5 also brings an adjustable effort setting to Anthropic’s Haiku line, letting users trade off cost and performance. On internal tests, it scored 72.4% against the offline subset of the OSWorld 2.1 computer-use benchmark. Haiku 4.5 scored 15.7%.
Anthropic pointed developers toward its larger models for more demanding work. Sonnet 5.5 and Opus 5.5 performed better on Terminal-Bench 4.0, where Haiku 5.5 scored 39.2% versus 70.6% for Sonnet 5.5, it said.
Anthropic also cut the price of Sonnet 5.5 cache reads by 50%, to $0.10 per million tokens from $0.20. The company said that lowers the cost of Sonnet 5.5 on most agentic tasks by about 20%.