Anthropic Releases Lower-Priced Claude Opus 5.5 With a Fable-Level Performance Claim

Anthropic has released Claude Opus 5.5 with a lower price than its predecessor and a performance pitch aimed at businesses watching AI costs. The company says the new model handles most tasks at the level of its more expensive Fable 5.1 model. That claim makes the price cut attractive, but it does not promise the same result on every job.

What the lower rate buys

Opus 5.5 costs $4 per million input tokens and $20 per million output tokens. Tokens are units of data used to process an AI request. Anthropic says the model is about 40% cheaper than its predecessor, Opus 5. That comparison is between Opus versions, not between Opus 5.5 and Fable 5.1.

Fable 5.1 costs $10 per million input tokens and $50 per million output tokens. Anthropic says Opus 5.5 can handle most tasks at Fable’s level, making the cheaper model a potential substitute for some work. A finished task’s bill also depends on how many tokens it uses; the listed rates alone cannot tell a customer its total cost.

Price competes with control

Businesses are sensitive to the expense of running large AI models as they use them for more tasks. OpenAI has also reduced its model prices. A lower rate could therefore help Anthropic compete for work where repeated use makes costs add up, even if a buyer still needs to judge the model’s results on its own tasks.

Open-weight models offer a different bargain: customers can customize them and run them on their own hardware. Moneycontrol, citing data from cloud application company Vercel, says Chinese models of this kind are gaining market share. A cheaper Claude rate addresses usage cost; for a buyer who wants to run a model on its own equipment, price is only part of the choice.

Safeguards alongside the cost pitch

Anthropic says Opus 5.5 underwent alignment testing and external evaluations before release. It also says the model has safeguards for high-risk cybersecurity and biology work. Those protections are part of the product a customer receives, alongside the lower rate; price and a broad performance claim cannot describe how it will respond to every sensitive request.

The release follows Anthropic chief executive Dario Amodei’s call for AI companies to slow frontier development to allow more safety and security testing. Opus 5.5 puts Anthropic’s two public priorities together: a cheaper model for customers and safeguards for higher-risk work. Buyers will have to assess both when deciding where to use it.