Anthropic launches Claude Opus 5.5 with faster output and lower costs

Anthropic has released Claude Opus 5.5, 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 than Opus 5 to run on typical workloads.

The launch brings changes for developers and subscribers. Opus 5.5 has lower API rates and faster output, while paid Claude plans are getting higher usage limits. Anthropic also says the model is stronger at long-running coding and computer-use tasks and communicates more clearly.

Opus 5.5 focuses on coding and long-running work

Anthropic says Opus 5.5 leads Opus 5 on its benchmarks for coding, computer use, and knowledge work. These are tasks where a model may need to work through several steps and use tools. One early tester completed a 680,000-line code migration in less than a day, according to Anthropic’s launch announcement. That is a company-selected example, not an independent test.

The company also says Opus 5.5 is more resistant to prompt-injection attacks, which try to make a model ignore its instructions, and less likely than recent models to take hard-to-reverse actions. Anthropic says its writing is clearer, puts key information up front, and is easier to follow during long work sessions. These are Anthropic’s own model and safety claims.

For developers, the Claude Platform model page lists a 1-million-token context window, a maximum output of 128,000 tokens, and the API model ID claude-opus-5-5.

API token prices are down 20%, while typical workload costs fall 40%

The 40% figure refers to total costs on typical workloads, not a 40% cut in each token price. Claude Opus 5.5 API rates are $4 per million input tokens and $20 per million output tokens, 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 lower prices combine with fewer tokens per task to reduce typical workload costs. It also says Opus 5.5 generates output more than 30% faster than Opus 5. The savings will vary by task and API usage; Anthropic’s 40% estimate is not a discount on a Claude subscription.

Subscribers get higher five-hour usage limits

Anthropic is increasing the five-hour usage limits for Pro, Max, Team, and seat-based Enterprise plans. It is also giving subscribers a rate-limit reset they can save and use later when they need it. These account limits are separate from the API prices, so the model’s lower token rates do not mean Claude subscription fees are dropping.

Opus 5.5 is available across Claude and cloud platforms

Users can select Opus 5.5 in Claude’s model picker. Developers can access it through the Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, and Claude Platform on AWS. Access and pricing can vary by provider.

The launch follows Claude Fable 5.1, Anthropic’s previous model for long-running coding and knowledge work. Our Claude icons guide shows where to find the model picker in the app. Anthropic says Claude Sonnet 5.5 and Haiku 5.5 will follow in the coming weeks.