Anthropic Unveils Fable: A More Affordable, Less Restrictive AI Model

Anthropic launched Fable and Mythos 5.1 on Tuesday, offering paired iterations of its latest AI technology. Beyond performance enhancements, the updated Fable version reduces token usage and tightens safety filters to minimize false-positive restrictions.
Like its predecessor, Mythos 5.1 is restricted to verified Anthropic partners focused on cybersecurity or life sciences. The unrestricted Fable 5.1 is now accessible via cloud platforms and the Anthropic API.
A key update is the adoption of zero data retention, enabling clients to deploy Anthropic models on private infrastructure without data exfiltration. Previously blocked for security reasons, this high-privacy Enterprise Frontier Safeguards service launches in June, allowing clients to manage monitoring protocols while the system continues to detect misuse by agents or users.
Anthropic confirmed that customer data remains secure, stating, “Anthropic has never trained on enterprise data without explicit permission, and never will.”
Consistent with past releases, the new models achieved top scores in benchmarks like Terminal-Bench 4.0 (CLI coding) and Humanity’s Last Exam (general reasoning). Anthropic also unveiled three pre-release scientific outputs, including a custom GPU optimization and a high-resolution Venus map compiled from existing imagery.
Accompanying the release is a detailed system card outlining model capabilities. It classifies Mythos as “low-risk” regarding autonomous AI development, noting that its capacity to accelerate internal AI research aligns with current industry trends.
Regarding general misbehavior, Mythos shows a slight increase in risk compared to Opus, likely due to its enhanced capabilities.
The system card notes, “Mythos 5.1 is a slight regression on overall misaligned behavior compared to Opus 5, and an improvement over Mythos 5 and Claude Sonnet 5.” It adds that while the model may cooperate with misuse or accept unverifiable authorization claims more readily than Opus 5, it is less prone to ignoring constraints, hallucinating inputs, or falsely claiming task completion than earlier versions.
Related article
Anthropic Enters AI Legal Tech Market as Competition Intensifies
Anthropic unveiled a suite of new chatbot capabilities on Tuesday, aimed at delivering automated support to legal practices. These enhancements expand upon Claude for Legal, the firm-specific platform introduced earlier this year, by adding specializ
OpenAI Closes Gap With Anthropic Among Business Users, New Data Shows
With OpenAI and Anthropic still distant from their anticipated IPOs and the release of detailed financial reports, we must turn to alternative indicators to gauge their business performance. Ramp, a corporate credit card and expense management platfo
Anthropic launches Opus 4.8 featuring new dynamic workflow tool
Anthropic unveiled Opus 4.8 on Thursday, marking the latest iteration of its premier public model. Priced identically to its predecessor, this update is now accessible across all platforms.Releasing just 41 days after Opus 4.7, Anthropic has accelera
Related Special Topic Recommendations
Comments (0)
0/500

Anthropic launched Fable and Mythos 5.1 on Tuesday, offering paired iterations of its latest AI technology. Beyond performance enhancements, the updated Fable version reduces token usage and tightens safety filters to minimize false-positive restrictions.
Like its predecessor, Mythos 5.1 is restricted to verified Anthropic partners focused on cybersecurity or life sciences. The unrestricted Fable 5.1 is now accessible via cloud platforms and the Anthropic API.
A key update is the adoption of zero data retention, enabling clients to deploy Anthropic models on private infrastructure without data exfiltration. Previously blocked for security reasons, this high-privacy Enterprise Frontier Safeguards service launches in June, allowing clients to manage monitoring protocols while the system continues to detect misuse by agents or users.
Anthropic confirmed that customer data remains secure, stating, “Anthropic has never trained on enterprise data without explicit permission, and never will.”
Consistent with past releases, the new models achieved top scores in benchmarks like Terminal-Bench 4.0 (CLI coding) and Humanity’s Last Exam (general reasoning). Anthropic also unveiled three pre-release scientific outputs, including a custom GPU optimization and a high-resolution Venus map compiled from existing imagery.
Accompanying the release is a detailed system card outlining model capabilities. It classifies Mythos as “low-risk” regarding autonomous AI development, noting that its capacity to accelerate internal AI research aligns with current industry trends.
Regarding general misbehavior, Mythos shows a slight increase in risk compared to Opus, likely due to its enhanced capabilities.
The system card notes, “Mythos 5.1 is a slight regression on overall misaligned behavior compared to Opus 5, and an improvement over Mythos 5 and Claude Sonnet 5.” It adds that while the model may cooperate with misuse or accept unverifiable authorization claims more readily than Opus 5, it is less prone to ignoring constraints, hallucinating inputs, or falsely claiming task completion than earlier versions.
Anthropic Enters AI Legal Tech Market as Competition Intensifies
Anthropic unveiled a suite of new chatbot capabilities on Tuesday, aimed at delivering automated support to legal practices. These enhancements expand upon Claude for Legal, the firm-specific platform introduced earlier this year, by adding specializ
OpenAI Closes Gap With Anthropic Among Business Users, New Data Shows
With OpenAI and Anthropic still distant from their anticipated IPOs and the release of detailed financial reports, we must turn to alternative indicators to gauge their business performance. Ramp, a corporate credit card and expense management platfo
Anthropic launches Opus 4.8 featuring new dynamic workflow tool
Anthropic unveiled Opus 4.8 on Thursday, marking the latest iteration of its premier public model. Priced identically to its predecessor, this update is now accessible across all platforms.Releasing just 41 days after Opus 4.7, Anthropic has accelera





Home






