Summary
Anthropic says Claude Opus 5.5 matches Claude Fable 5.1 on most work and costs 40% less than Opus 5 on typical workloads. The new model also has lower token prices and is designed for complex, multi-step tasks.
Anthropic introduced Claude Opus 5.5 on September 22, 2026, describing it as the first model in its Claude 5.5 family. The company says it performs at the level of Claude Fable 5.1 on most work while costing 40% less to run than Opus 5 on typical workloads. It also says Opus 5.5 generates output more than 30% faster than its predecessor.
Performance in longer tasks
Anthropic presents Opus 5.5 as an upgrade for complex work that involves multiple steps and tool use, particularly coding. In one company-run test, Opus 5.5 and Fable 5.1 each translated HAProxy—a widely used program for balancing web traffic across servers—from C into Rust. Both rewrites passed nearly all of HAProxy’s regression tests. Opus 5.5 finished in 9.5 hours, compared with 12 hours for Fable 5.1, and Anthropic says its run cost 51% less.
The announcement also reports results across coding, computer-use and knowledge-work benchmarks. These tests assess tasks such as completing command-line work, handling computer interfaces and carrying out professional workflows. Anthropic cautions that benchmark margins have become a less reliable guide to real-world differences at this level of capability; in its own use, it says the gap between Opus 5.5 and Fable 5.1 is narrower than the scores suggest. Some comparisons also use different effort settings, so the results should be read in the context of each test’s setup.
For software teams, the cost claim reflects more than the price of an individual token: Anthropic says Opus 5.5 uses fewer tokens per task as well as costing less per token. That can matter when an agent runs for a long time or makes many calls while working through a large codebase. The company cites an early tester that completed a 680,000-line code migration in less than a day, work the tester estimated would take an engineering team weeks; this is a reported example, rather than a general measure of performance.
Pricing and safeguards
For every million tokens, Anthropic lists Opus 5.5 input at $4 and output at $20, down from $5 and $25 for Opus 5. Cache reads are $0.20 per million tokens, compared with $0.50 for Opus 5. Anthropic says its tests found a 40% cost reduction on typical workloads at default settings. A faster mode in Claude Code and the Claude Platform is listed at up to 2.5 times the speed, priced at $8 per million input tokens and $40 per million output tokens.
Anthropic says Opus 5.5 achieved the strongest result to date on its automated behavioral audit, which tests Claude models across thousands of simulated scenarios. The company says the model is less likely than recent models to take hard-to-reverse actions or exceed its assigned boundaries, and is more resistant to prompt injection than Opus 5. It reports that external evaluators, including Frontier Design and METR, tested the model before release.
The company is applying additional access controls to biology and cybersecurity uses. Vetted organisations can apply to its Life Sciences Verification Program to use Opus 5.5 for biology research. Anthropic says it plans to expand its Cyber Verification Program in the coming weeks so verified cybersecurity practitioners can use the model. It is also increasing five-hour usage limits on Pro, Max, Team and seat-based Enterprise plans.