Sonnet 5.5 is the second model in Anthropic's Claude 5.5 family, six days after Opus 5.5. It is aimed at well-scoped everyday work — bug fixes, documents, slides, spreadsheets — and is the first Sonnet to ship with the cyber safeguards used on Anthropic's most capable models.
Published · 2 sources
Confirmed. The release, the unchanged token prices, the speed and cost claims, and the cyber-safeguard change are in Anthropic's own announcement, with the price and the everyday-work positioning also reported by Reuters. A pricing-page correction, or a published Haiku 5.5 date, would be what would move this entry.
If you use Sonnet for ordinary coding and document work, the sticker price did not change: $2 per million input tokens, $10 per million output tokens, and $0.20 per million cache reads. Anthropic's claim is that the same job now costs up to 30% less because the model uses fewer tokens, and that it writes output more than 30% faster than Sonnet 5.
Cyber safeguards that used to sit on the larger models now apply here. Biology safeguards are unchanged from Sonnet 5. Haiku 5.5, the cheap high-volume model, is still not out.
Anthropic
anthropic.com
Primary source
Open original (opens in a new tab)Reuters
reuters.com
Secondary
Open original (opens in a new tab)Learn the idea behind this story
A mid-tier model that finishes scoped tasks with fewer tokens changes when it is worth stepping up to a larger model — the same cost-and-fit judgment the lab practises.
Open the prompt lab