Anthropic released Claude Haiku 5.5 as its fastest and most capable small model. It targets high-volume, cost-sensitive tasks like summarization and subagents. The model costs about 75% less than Haiku 4.5 on average. It also introduces an adjustable effort setting for cost or intelligence trade-offs. Sonnet 5.5 cache reads are now half price, cutting its cost by around 20% on agentic work. Max and Team subscribers receive new monthly API credits of $100 to $500. Haiku 5.5 is available now on AWS, Google Cloud, Azure, and the Claude Platform. Its safeguards are more restrictive than Haiku 4.5 but less so than Sonnet 5.5. It performs strongly on knowledge, computer use, and coding benchmarks. Early customers report faster inference and lower latency in production use.
What changed
Haiku 4.5 was the prior small model with higher cost and lower capability.
What it unlocks
Running high-volume, cost-sensitive tasks like summarization, subagents, and browser use at much lower cost and higher speed.
- 75% cheaper than Haiku 4.5 on average
- 90% of requests ≤100k tokens
- 20% cheaper Sonnet 5.5 on agentic work
- $100–$500 monthly API credits for Max/Team
What you need to act on it
- Claude API access
- Max or Team plan for new credits
Sources