Anthropic

Anthropic releases Claude Haiku 5.5

Anthropic has released Claude Haiku 5.5, the third model in the Claude 5.5 family and the first Haiku model with adjustable reasoning effort. The launch comes nearly a year after Claude Haiku 4.5.According to Anthropic, Haiku 5.5 is its fastest and most capable small model to date, designed for high-volume, cost-sensitive workloads.
Anthropic releases Claude Haiku 5.5
Anthropic has released Claude Haiku 5.5, the third model in the Claude 5.5 family and the first Haiku model with adjustable reasoning effort. The launch comes nearly a year after Claude Haiku 4.5.According to Anthropic, Haiku 5.5 is its fastest and most capable small model to date, designed for high-volume, cost-sensitive workloads.
Anthropic releases Claude Haiku 5.5
The model is particularly well-suited to classification, summarization, data extraction, browser automation, real-time assistants, and other repetitive tasks. It can also serve as a subagent, handling focused tasks delegated by larger models such as Claude Sonnet and Opus.

API Pricing: Up to 90% Lower Than Haiku 4.5

Claude Haiku 5.5 introduces two pricing tiers based on prompt length.
Anthropic releases Claude Haiku 5.5
For shorter prompts, these rates match the standard API pricing of OpenAI's GPT-6 Luna.According to , Haiku 5.5 is 90% cheaper than Haiku 4.5 for prompts of up to 100,000 tokens. These shorter prompts accounted for approximately 90% of requests to the previous model.For longer prompts, token prices are 50% lower. Across typical workloads, Anthropic estimates that the average cost of running Haiku 5.5 is approximately 75% lower than Haiku 4.5, accounting for differences in token usage.

Performance and Adjustable Reasoning

Claude Haiku 5.5 is the first Haiku model to support adjustable reasoning effort, allowing developers to balance cost and reasoning performance depending on the task.Anthropic also reports significant improvements in computer use. On the OSWorld 2.1 benchmark (offline subset), Haiku 5.5 scored 72.4%, compared with 15.7% for Haiku 4.5.The model supports a context window of up to 1 million tokens and a maximum output of 128,000 tokens.Haiku 5.5 is primarily designed for frequent, well-defined tasks where using more powerful models such as Sonnet or Opus would be unnecessarily expensive. For complex coding and reasoning tasks, Anthropic continues to recommend its larger models.

Availability

Claude Haiku 5.5 is available through the Claude API under the model identifier claude-haiku-5-5, as well as through supported cloud platforms, including AWS, Google Cloud, and Microsoft Azure.The model is also available through , where users can access it alongside other AI models through a unified platform.
Anthropic releases Claude Haiku 5.5 :: BotHub in Russia