The model ID is claude-haiku-5-5; the page lists availability through the Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, and Claude Platform on AWS.
The documented limits are a 1M-token context window and 128K maximum output. The Batch API beta supports up to 300K output tokens with the output-300k-2026-03-24 beta header. Haiku 5.5 has adaptive thinking, a default medium effort setting, and a June 2026 knowledge cutoff.
For prompts up to 100K tokens, listed input and output prices are $0.10 and $0.50 per million tokens. Above that threshold they are $0.50 and $2.50. Batch API input and output receive a 50% discount; cache reads are $0.01 per million tokens below the threshold. The page says retirement will be no sooner than October 7, 2027. It does not establish comparative quality, measured latency, or practical performance for a particular workload.