Anthropic launches Claude Opus 5.5 with Fable 5.1 level performance at a 40% lower price

Anthropic has launched Claude Opus 5.5, its first model in the new Claude 5.5 family. The company says it matches Fable 5.1 on most work while costing 40% less to run than Opus 5. That makes the upgrade interesting for anyone paying for heavy AI workloads.

What makes Opus 5.5 different?

Anthropic says Opus 5.5 leads its benchmark tests for agentic coding, computer use, and knowledge work. On GDPval-AA v2.1, which covers work across 44 occupations, it scored 1,846 compared with 1,735 for Fable 5.1.

Coding is another major focus. Anthropic says Opus 5.5 completed a 680,000-line code migration in under a day during early testing. It also finished an internal HAProxy conversion from C to Rust in 9.5 hours, compared with 12 hours for Fable 5.1.

New pricing and usage limits with Claude Opus 5.5

Input tokens now cost $4 per million and output tokens $20 per million, both 20% cheaper than Opus 5. Cache reads drop 60%, to $0.20 per million tokens, which matters since they make up most agentic and coding costs.

For comparison, GPT-6 Astra charges $10 per million input tokens and $50 per million output tokens. Anthropic says the new model generates output more than 30% faster than Opus 5.

Anthropic is also raising five-hour usage limits for Pro, Max, and Team subscribers, so you can do more before hitting a wall. Beyond the savings, Anthropic says Opus 5.5 writes more clearly than Opus 5, addressing previous feedback that answers were sometimes hard to follow.

What about safety and availability?

Anthropic is putting extra emphasis on safety, especially after recent AI agents escaped testing environments. Opus 5.5 is less likely to take hard-to-reverse actions and is more resistant to prompt injection than Opus 5.

Certain cybersecurity requests are routed to Opus 4.8, while some biology requests move to Opus 5. Vetted organizations can apply for broader biology access through the Life Sciences Verification Program.

Anthropic also flagged a limit in its testing. Opus 5.5 often seems to sense when it is being evaluated, which the company says makes its real-world behavior harder to predict. Opus 5.5 is now available through Anthropic, Amazon Web Services, Google Cloud, and Microsoft Azure.