Anthropic released Claude Fable 5.1 and Claude Mythos 5.1, two versions of its same frontier AI model with different safeguards, claimed to “excel” at coding and knowledge work, with Fable 5.1 available in general and Mythos 5.1 restricted to trusted access programs for cybersecurity and life sciences experts.
In a post on Tuesday, Anthropic said that Claude Fable 5.1 is capable of much higher performance than its predecessor Claude Fable 5 and when set to “low or medium effort”, its result is similar to Fable 5 but at a much lower cost.
“Claude Fable 5.1 sets a new standard for coding, knowledge work, and long-running problem-solving tasks. avoids shortcuts that result in poorer-quality work, and it’s smart enough to fix the root causes of software issues. For example, in testing by the investment firm Millennium, Fable 5.1 found the cause of a rare crash in its internal systems that none of its engineers (or any other model) had been able to explain after several years of trying,” claimed Anthropic.
We’re introducing Claude Fable 5.1 and Claude Mythos 5.1.
They’re the world’s most advanced models for coding and knowledge work. pic.twitter.com/8P9PSrWPi3
— Claude (@claudeai) September 1, 2026
Anthropic said that the two models are built with different levels of safeguards and also announced the launch of Enterprise Frontier Safeguards (EFS) offering data privacy to users with zero data retention policy.
According to Anthropic EFS will work by storing data in cloud infrastructure controlled entirely by the customer, not by the company. It will be made available to enterprise customers in phases, beginning later this fall. Until EFS is available, eligible customers will be able to use Fable 5.1 with zero data retention.
It further stated that Fable 5.1 will cost about 25 percent less than Fable 5 for typical workloads where usage is billed by token, following a reduction in the price of cache reads. For highly agentic workloads, the company expects savings of up to about 45 percent.
Anthropic has also updated its safeguards for cybersecurity and biology to reduce false positives. The company said its latest cybersecurity safeguards block 60 percent fewer false positives than previous safeguards.
“Fable 5.1 can now be used to discover software vulnerabilities—though not to develop exploits for them. In biology, we’ve established an access program, developed in partnership with the US government, to enable access to Claude Mythos 5.1’s advanced biology capabilities. We expect to open enrollment for scientists soon,” the company said.
Anthropic highlighted that testing revealed that Mythos 5.1 has stronger scientific research capabilities than its predecessors. It also demonstrated the strongest cybersecurity capabilities of any model it has released when its cybersecurity safeguards were disabled. However, the company said its evaluations placed the model within the lower category of risk under its frontier compliance framework.
The company also tested Mythos 5.1 against chemical and biological risks, including its potential to assist in creating chemical or biological weapons. Anthropic said the model’s capabilities were greater than those of Mythos 5, but remained below the new risk tier under its Responsible Scaling Policy.
Anthropic’s alignment evaluations also found improvements compared with Mythos 5. The company said Mythos 5.1 was less likely to attempt to access resources outside its testing environment, justify unsafe actions through motivated reasoning or ignore explicit constraints in pursuit of a user’s goals.
However, Anthropic acknowledged that the model can still bypass some approval systems and automated classifiers. Alongside the new safeguards, Anthropic said Fable 5.1 includes stronger protections against model distillation, a process used to extract capabilities from advanced AI systems.
The company informed that Claude Mythos 5.1 is identical to Fable 5.1, but it offers more permissive safeguards for vetted individuals and organizations whose work is affected by the cybersecurity and life sciences restrictions . Mythos 5.1 will be available through two trusted access programs.
The first is cyber verification program, which will currently provide access to certain Opus- and Sonnet-class models with reduced cyber safeguards for defensive security work. In the near future, this program will also include access to Claude Mythos-class models.
The second is the life sciences verification program, which has been designed so that life sciences professionals can use Claude Mythos 5.1 with safeguards designed for professional research and development activities (while all other safeguards remain in place).
“In partnership with the US government, we have enrolled our first participants, and we plan to expand access to this program to the broader life sciences community,” the company said.
In June the company’s earlier models Claude Fable 5.0 and Claude Mythos 5.0 were banned by the US government for all except US nationals, citing national security concerns. However, a few weeks later, the AI company re-deployed the Mythos 5 model to a set of US organizations.
The government had partially lifted export control over Claude Mythos 5 model to organizations that operate and defend critical infrastructure. Later Anthropic had announced that its model Claude Fable 5 was also accessible to users worldwide.
Also Read: Anthropic Resumes External Testing Following July Claude Hacks
