On September 22, Anthropic announced Claude Opus 5.5, the first model in its new Claude 5.5 family. According to the company, Opus 5.5 delivers performance on par with Claude Fable 5.1 for most real-world work, while running typical Opus 5 workloads at roughly 40% lower cost.
The model is now available on the Claude Platform under the name claude-opus-5-5, and via Amazon Web Services, Google Cloud, and Microsoft Azure.
Performance and pricing
In Anthropic’s internal and early-customer tests, Opus 5.5:
- completed a 680,000-line code migration in under a day;
- audited and fixed a 200,000-line codebase in under 3 hours (Opus 5 needed over 20 hours and 2.5× more tokens);
- translated the HAProxy load balancer from C to Rust, finishing faster than Fable 5.1 at 51% lower cost.
Pricing for Opus 5.5 reflects lower serving compute:
- Input: $4 / 1M tokens (20% less than Opus 5)
- Output: $20 / 1M tokens
- Cache reads: $0.20 / 1M tokens (60% less than Opus 5)
A Fast mode is available in Claude Code and the Claude Platform, offering up to 2.5× higher speed at higher per-token rates.
Coding and agent use cases
Opus 5.5 is aimed squarely at agentic coding and large-scale codebase migrations and audits. On benchmarks like FrontierCode, Terminal Bench 4.0, and CursorBench, it matches or beats GPT-6 Astra and GPT-5.6 Sol while costing a fraction per task.
For enterprise agent deployments, the model includes:
- a classifier that screens every action before execution;
- an open-source sandbox that security teams can audit;
- integrated code review to catch vulnerabilities before merge.
Safety and alignment
Anthropic says Opus 5.5 achieves the best results so far on its automated behavioral audit, which evaluates Claude across thousands of simulated scenarios. The model is:
- more resistant to prompt injection than Opus 5;
- less likely to take hard-to-reverse actions or operate outside given boundaries;
- tested on longer tasks, impossible tasks, and scenarios modeled on real incidents.
Because its capabilities in biology and cybersecurity are comparable to Claude Mythos 5.1, Opus 5.5 ships with safeguards similar to those on Claude Fable 5.1. Vetted organizations can apply to Anthropic’s Life Sciences Verification Program and the expanding Cyber Verification Program for access to higher-risk use cases.
Anthropic plans to follow with Claude Sonnet 5.5 and Claude Haiku 5.5, bringing similar gains in performance, efficiency, and safety to the rest of the lineup.
Source: Anthropic
