Anthropic's Claude Fable 5.1, Mythos 5.1 launch addresses price and data retention concerns

Anthropic launched Claude Fable 5.1 and Claude Mythos 5.1 and moved to address two big issues: Zero data retention and price.

The company billed Fable 5.1 and Mythos 5.1 as big advances for coding and knowledge work, but the big win for customers is on zero data retention and price. Anthropic caught flack for a data retention policy in June along with the Claude Fable 5 and Mythos 5 launch. The new policy meant Anthropic could require a 30-day retention for all traffic on models to defend against misuse and cyberattacks.

Customers pushed back and Anthropic launched Enterprise Frontier Safeguards (EFS), which consists of a system that "gives customers complete privacy (the same as a zero data retention policy) while still being state-of-the-art at preventing adversarial use. EFS works by storing data in cloud infrastructure controlled entirely by the customer, not Anthropic."

EFS will be available to enterprise customers beginning this fall.

The zero data retention issue was a big unforced error by Anthropic--especially when OpenAI and rivals stuck with zero data retention. Box CEO Aaron Levie recently addressed the issue on an earnings conference call.

"My guess realistically is Anthropic will evolve their stance on this because they'll see it in the revenue. And they'll have to change course. And I think we'll just see more and more modern approaches to how these companies will both meet their safety -- internal kind of AI safety requirements with offering things that kind of technically resembles ZDR for their users."

Price has also been a big issue and Anthropic is trying to fend off open models as well as Google, OpenAI, SpaceX and Meta all forcing prices lower.

• The most important LLM benchmark today: Price
• Constellation ShortList™ Frontier Large Language Models
• Palantir: Worst nightmare for OpenAI, Anthropic?
• SpaceXAI, Meta puts pricing squeeze on Anthropic, OpenAI

Anthropic said:

"Fable 5.1 will cost an estimated 25% less than Fable 5 for typical workloads, wherever usage is billed by token. This is because we’re reducing our pricing on cache reads (where the model reads inputs that have already been processed and stored). For highly agentic work, the savings will often be much larger—up to approximately 45%."

It's notable that Anthropic addressed customer feedback issues before getting to Fable 5.1 and Mythos 5.1 performance. The model handily tops Anthropic's predecessor models as well as OpenAI's GPT-5.6 Sol.

Mythos 5.1 has more permissive safeguards for vetted individuals and organizations working on cybersecurity and life sciences.