Claude Haiku 5.5 — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Anthropic launched Claude Haiku 5.5, calling it its cheapest, fastest, and most capable small model, and made it available immediately on AWS, Google Cloud, and Microsoft Azure.
- Claude Haiku 5.5 costs roughly 75% less to run than Haiku 4.5 on average, with the source noting it is 90% cheaper for requests up to 100,000 tokens — a category that made up 90% of prior Haiku requests.
- Claude Haiku 5.5 is the first Haiku-class model to ship with an adjustable effort setting, letting users tune the model to optimize for cost or intelligence.
- Claude Sonnet 5.5 cache read prices were cut 50% (from $0.20 to $0.10 per million tokens), reducing the cost of most agentic tasks by roughly 20%.
- Anthropic introduced new monthly API credits: $100/month for Max 5x users, $200 for Max 20x, and up to $500 pooled for Team subscribers, usable on any of its models.
- Haiku 5.5's cybersecurity safeguards are more restrictive than Haiku 4.5's but less restrictive than those applied to other recent Claude models; biology safeguards match those of Sonnet 5, Sonnet 5.5, and Opus 5.
- Anthropic updated its Python and TypeScript SDKs to add computer use and browser use in beta, noting Haiku 5.5 is especially well-suited to these speed-sensitive tasks given its combination of speed, capability, and price.
Why it matters: With Haiku 5.5 priced 90% lower than Haiku 4.5 for the request size that accounts for 90% of prior Haiku usage, the cost calculus for routing repetitive and subagent work to Claude shifts sharply; a 50% cut to Sonnet 5.5 cache reads on top means Anthropic is now competing on volume pricing, not just flagship capability — and developers running high-throughput workloads stand to gain the most.
Ask SkimNews
