GMAsia
    AI News·7 Oct 2026·via Testing Catalog

    Anthropic launches Claude Haiku 5.5 at lower API prices

    Anthropic has launched Claude Haiku 5.5, its smallest and most cost-effective model, designed for high-volume AI tasks, according to the company. The release also includes reduced cache-read costs for Sonnet 5.5 and new monthly API credits for Max and Team subscribers.

    Nexa's Summary

    Haiku 5.5 is positioned as Anthropic's most economical small model, targeting applications like summarization, classification, and customer support. Its pricing structure, which includes lower token costs for prompts up to 100,000 tokens and a stated 75% reduction in average running costs compared to its predecessor, suggests a focus on making AI accessible for repetitive, high-volume operations. This model can also function as a subagent, indicating its role within a broader AI architecture.

    A notable feature is the adjustable effort setting, allowing developers to balance cost against intelligence for specific tasks. This flexibility could be important for enterprises managing large-scale AI deployments where efficiency gains from marginal performance adjustments can lead to substantial cost savings. The model's availability across major cloud platforms like Amazon Web Services, Google Cloud, and Microsoft Azure also broadens its potential reach.

    The performance metrics provided, such as significant score increases on OSWorld 2.1 and Terminal-Bench 4.0 compared to Haiku 4.5, suggest a substantial improvement in capability for its intended use cases. While Sonnet 5.5 and Opus 5.5 remain designated for more complex agentic coding, Haiku 5.5 aims to handle narrower, repeatable jobs at scale. This differentiation in model roles indicates Anthropic's strategy to cater to a spectrum of AI application needs.

    Share this article

    Go deeper
    Original reporting by Testing CatalogWe don't republish, read the full story →

    Related reading

    6 stories