Claude Opus 5.5 delivers Fable 5.1 performance – and costs 40% less
Anthropic’s newest Claude model aims to make frontier AI cheaper, faster, and safer, while giving subscribers more breathing room before those frustrating usage limits kick in.
Claude Opus 5.5, the workhorse AI from Anthropic, now delivers Fable 5.1 performance for most tasks at about 40% lower costs. According to Yashodha Bhavnani, VP of AI Products at Box, Opus 5.5 uses a third of the tokens compared to Opus 5, and its answers are 40% less verbose without losing accuracy. Box expects this to benefit teams running agents across content in sectors like financial services and public sector.
Opus 5.5 is available now. Anthropic also announces the upcoming launch of Claude Sonnet 5.5 and Haiku 5.5. Opus 5.5 generates output 30% faster than Opus 5 and costs 20% less per token. Tokens are priced at 20% less than Opus 5, and subscription users receive a 20% increase in usage limits. Opus 5.5 also reportedly needs fewer tokens for higher quality work and communicates more naturally, with less obsequious behavior.
Anthropic emphasizes improvements in alignment testing and safeguards for high-risk areas, such as cybersecurity and biology, citing CEO Dario Amodei's emphasis on AI development safety. GitHub's chief product officer, Mario Rodriguez, reports that Opus 5.5 performs better than Opus 5 in real software work, using fewer tokens and steps on GitHub Copilot CLI and VS Code.
Deloitte Consulting LLP's CIO, Carl Bennett, shares that Opus 5.5 catches 72% of known bugs in code reviews compared to 56% for Opus 5, with fewer false alarms and lower output.
Written by urgent.news from ZDNet's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
