Claude Opus 5.5 is out: Fable-level results, lower price
Anthropic released Opus 5.5, the first model of its 5.5 family. It beats Fable 5.1 on agentic work and costs 20% less than Opus 5.
What changed?
Opus 5.5 launched on September 22 in Anthropic's own apps and on AWS, Google Cloud and Microsoft Azure on the same day. Anthropic says typical workloads cost 40% less than on Opus 5 and output is more than 30% faster. Pricing is $4 per million input tokens and $20 per million output; cache reads fell to $0.20.
Where it leads
In the tests Anthropic published, the model beats Fable 5.1 on agentic coding, knowledge work, computer use, chart reading and multidisciplinary reasoning. External evaluators such as METR tested it before release. Anthropic says Sonnet 5.5 and Haiku 5.5 are coming soon.
Why it matters
OpenAI shipped two new models the same day. Both companies are cutting prices, which means businesses running AI agents can do the same work for less.