Anthropic Launches Claude Opus 5.5: Fable-Level Performance at a 20% Lower Price

Anthropic officially launched its new large language model, Claude Opus 5.5, on September 22. The company described it as "a major step up from Opus 5 with improvements in performance and safety." Two things stand out: top-tier capability at a lower price, and a clear jump in measured coding and agentic performance.

Cheaper to run, stronger on tasks

Opus 5.5 is priced at $4 per million input tokens and $20 per million output tokens — each 20% below Opus 5's $5 and $25. Cache reads, which heavily shape cost in repeated calls, fall to $0.20 per million tokens, 60% cheaper than Opus 5's $0.50. Anthropic says that on typical workloads Opus 5.5 costs about 40% less to run than Opus 5, and generates output more than 30% faster.

Big gains on coding and agentic benchmarks

In benchmarks published by Anthropic, Opus 5.5 beat Opus 5 across coding, computer use, and knowledge work — with the largest gains on terminal and coding tests that mimic real development environments.

Anthropic claims Opus 5.5 matches or exceeds OpenAI's GPT-6 Astra on several benchmarks while costing just 20-40% as much. One early tester reportedly audited a 200,000-line codebase in under three hours — a task that took Opus 5 more than 20 hours — and another completed a 680,000-line code migration in less than a day.

Safety and communication upgrades too

Anthropic emphasized safety alongside raw performance. The company says Opus 5.5 earned "the best scores of any model to date" on its automated behavioral audit, and its prompt-injection resistance tied the higher-tier Fable 5.1 on the Gray Swan benchmark. New safeguards cover cybersecurity, biology, and distillation protection.

The model's communication style was also revised: Opus 5.5 puts the most important information up front, uses less jargon, and better follows user-provided writing guidelines.

Why it matters

Opus 5.5 blurs the old split between a pricier top model and a cheaper lower one, pushing top-tier capability into a lower price band. That strategy lines up with the frontier price war unfolding as rivals cut their own prices the same week. In a market where raw performance alone no longer locks in customers, this launch signals that price-performance is becoming the next battleground.