Anthropic has released Claude Opus 5.5, a new version of its high-end model that the company says is cheaper to run than Opus 5 while keeping much of the performance it associates with its Fable 5.1 model. The company announced the release on Sept. 22 and says Opus 5.5 is available in Claude, Claude Code and the Claude Platform.
The practical change for developers is pricing. Anthropic lists Opus 5.5 at $4 per million input tokens and $20 per million output tokens, down from $5 and $25 for Opus 5. Cache reads, which can account for much of an agentic coding workflow, are listed at $0.20 per million tokens rather than $0.50. Anthropic says the combination of lower prices and fewer tokens used per task cuts typical costs by 40% and that the model generates output more than 30% faster. Those are the company's own measurements, not an independent price-performance test.
A release with guarded access
Anthropic is also tying the release to controls for high-risk work. The company says Opus 5.5 is comparable to its more restricted Mythos 5.1 model in biology and cybersecurity, so it is shipping with safeguards similar to those used for Fable 5.1. Organizations can apply to the Life Sciences Verification Program for biology research, according to Anthropic. The company says it will expand its Cyber Verification Program in coming weeks so verified cybersecurity practitioners can use Opus 5.5 for that work.
The announcement also points to an action-screening classifier, an open-source sandbox and code review as parts of the product's security approach. Anthropic says the model is less likely than recent releases to take hard-to-reverse actions or operate beyond its assigned boundaries, and that it performs better than Opus 5 against prompt injection in its tests. Those safety and benchmark claims come from Anthropic's evaluation program; the company notes that the model still has limits and that benchmark margins do not always translate cleanly into real-world differences.
For subscribers, Anthropic says it is raising five-hour usage limits for Pro, Max, Team and seat-based Enterprise plans and offering a rate-limit reset that subscribers can save for later. Sonnet 5.5 and Haiku 5.5 are planned for the coming weeks. The fuller technical and pricing details are in Anthropic's release announcement.
Tech