Anthropic launches Fable 5.1: Up to 75% price reduction, enhancing programming and research capabilities.

Anthropic released a new version of its flagship AI model, Fable 5.1, on Tuesday. While significantly reducing some usage costs, it also enhances programming and research capabilities. This move directly responds to price pressure from enterprise customers and reflects the increasingly fierce market competition among US AI model vendors.

The most significant change in the new version is the adjustment to the pricing strategy: Anthropic has reduced its fees by 75% for the process of retrieving processed information from the model, which the company says will result in a "significant decrease" in overall operating costs.

Dianne Penn, the company's product management manager, pointed out that this stage accounts for half or even more of the token consumption in some complex and long tasks, and is one of the most frequently cited pain points by customers.

At the same time, Anthropic also announced that it will allow enterprise customers to retain data on their own cloud infrastructure while still being subject to Anthropic's security reviews. This policy marks a significant shift in the company's previous data retention policies.

In addition, the invitation-only advanced model Mythos 5.1 will also be available to a limited number of pre-selected users on the same day.

Price reduction logic: responding to cost pressures and competitive landscape

Fable 5.1 maintains the same pricing structure for new input tokens as its predecessor, Fable 5. The price adjustment is mainly focused on the "cache reads" part—that is, the billing process when the model rereads previously processed content, with a reduction of up to 75%.

Dianne Penn stated that this step often consumes more than half of the token usage in tasks such as software development and code review that require the model to traverse the entire application codebase. "This is an area we hope to improve, and it's also in response to customer feedback," she said.

In addition to the price reduction, the new model itself has also been optimized to accomplish the same task with fewer tokens.

The context for this adjustment is clear: competition in the model market is intensifying; at the same time, enterprise buyers are becoming more cautious about AI spending and urgently need a more predictable cost structure.

Capability Upgrade: Breakthroughs in Both Programming and Scientific Research Scenarios

In terms of performance, Fable 5.1 has focused on enhancing two types of professional scenarios.

In programming, the new model excels at handling long-chain tasks spanning the entire codebase, including large-scale software project development and systematic code review. In scientific research, the model shows improved performance in tasks such as experimental design, experimental process simulation, and reading high-density graph data.

The security classifier has also been optimized accordingly—the false positive rate has decreased significantly. This means that users working in cybersecurity and biology-related fields will experience fewer false positives, potentially improving operational efficiency.

It is worth noting that the U.S. government had previously imposed access restrictions on the Fable and Mythos models in June due to security concerns. The restrictions were subsequently lifted after Fable 5 added a classifier capable of identifying and blocking more cybersecurity tasks; access to Mythos 5 was later restored, but to a lesser extent.

Data policy shift: Corporate sovereignty demands are being addressed

Anthropic's announced data retention policy adjustments are perhaps one of the most attractive changes for enterprise customers in this update.

The new policy allows enterprise users to keep data on their own cloud computing infrastructure when using Anthropic's most powerful capability model, rather than being forced to store it on Anthropic servers—while Anthropic will still perform security checks on the relevant data.

The original intention of the policy was to reduce cybersecurity risks, but this also deterred some enterprise customers who have strict requirements for data sovereignty.

With IPO imminent, revenue expansion becomes key to valuation support.

The release of Fable 5.1 is part of a larger strategy by which Anthropic and OpenAI are competing to expand their enterprise customer base and cover professional scenarios such as programming, cybersecurity, financial services, science, and healthcare.

Anthropic projects annualized revenue to exceed $65 billion, representing more than seven times the growth rate seen at the end of 2025.

Both Anthropic and OpenAI have submitted private listing applications, with Anthropic expected to go public as early as this fall.

With the IPO window approaching, it is crucial for both companies to continue expanding their market share and converting their high valuations into visible revenue growth.

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.