Anthropic releases Claude Haiku 5.5 at about 75 percent lower running cost than Haiku 4.5, and cuts Sonnet 5.5 cache prices

Anthropic has released Claude Haiku 5.5 and calls it its cheapest, fastest and most capable small model. The company says it is built for high-volume, cost-sensitive work such as summaries, compactions, database queries and classification. It is also meant to work as a subagent alongside Opus 5.5 and Sonnet 5.5 on coding tasks, and its speed suits live customer support and browser use. Anthropic says Haiku 5.5 costs about 75 percent less to run on average than Haiku 4.5. In its published benchmarks, Haiku 5.5 scores 72.4 percent on an offline subset of the OSWorld 2.1 computer-use test, compared with 15.7 percent for Haiku 4.5, 48.9 percent for OpenAI's GPT-6 Luna and 83.9 percent for Sonnet 5.5. On the GDPval-AA v2.1 knowledge-work test, it scores 1620, compared with 735 for Haiku 4.5, 1437 for GPT-6 Luna and 1840 for Sonnet 5.5. Anthropic is also halving the price of Sonnet 5.5 cache reads, which it says makes that model about 20 percent cheaper on most agentic work. Claude Max and Team subscribers get a new monthly API credit for building agents and applications on the Claude Platform. The release came on the same day OpenAI launched GPT-6 in ChatGPT, and it lowers costs for developers who run many cheap model calls in agent pipelines.