Photo credit: X/@AnthropicAIAnthropic has introduced Claude Fable 5, a new artificial intelligence model that marks the first public release of the company's powerful Mythos-class technology. Announced on June 9, Fable 5 is built on the same foundation as Claude Mythos 5, a model that had previously been restricted due to concerns about its advanced cybersecurity capabilities. The company says Fable 5 is its most capable generally available AI model yet, delivering improvements in software engineering, scientific research, visual understanding, reasoning and long-duration autonomous tasks. Unlike Mythos 5, however, Fable 5 comes with a range of safeguards designed to prevent misuse in sensitive areas such as cybersecurity, biology and chemistry. The launch represents Anthropic's attempt to balance frontier AI performance with responsible deployment. Key takeawaysClaude Fable 5 is Anthropic's first publicly available Mythos-class AI model.It shares the same core architecture as Claude Mythos 5.Anthropic has added safety restrictions for broader public use.The model excels in coding, research, vision and long-horizon reasoning tasks.High-risk requests in cybersecurity, biology and chemistry are routed to older models.Mythos 5 remains available only to selected organisations through Project Glasswing.Fable 5 is available through Claude and API services globally. Claude Fable 5 explainedClaude Fable 5 is a Mythos-class AI model developed by Anthropic and designed for general availability. The company describes it as the most capable model it has released to the public, outperforming previous Claude models in software engineering, scientific research, visual reasoning and complex knowledge work. Anthropic says the model performs particularly well on longer and more demanding tasks, where it can sustain reasoning over extended periods. Fable 5 is multimodal, allowing it to process text and visual information while supporting enterprise and developer use cases through APIs and Claude subscriptions. While it retains the underlying capabilities of Mythos 5, Anthropic has added controls that restrict access to sensitive functions, enabling broader deployment without exposing the full power of the unrestricted model. FeatureDetailsFoundationBuilt on Mythos 5 architectureCodingAdvanced software engineering capabilitiesMultimodalSupports text and visual understandingResearchStrong performance in scientific and knowledge tasksLong-horizon reasoningHandles complex, extended workflowsEnterprise supportAvailable through API and enterprise plansSafety controlsRestrictions on high-risk domainsVision capabilitiesEnhanced image and document analysisGlobal accessAvailable through Claude ecosystemAutonomous workflowsCan work on tasks for extended durationsWhy anthropic kept mythos under restricted access Before Fable 5's launch, Anthropic limited access to Mythos because of concerns about the model's extraordinary capabilities, particularly in cybersecurity. The company had previously indicated that Mythos could identify software vulnerabilities at a level that raised concerns among governments, infrastructure operators and security experts. As a result, access was restricted to a small group of vetted organisations through Project Glasswing. Anthropic argued that releasing the model without safeguards could create risks if malicious actors used its capabilities to discover or exploit vulnerabilities.The company spent months evaluating potential threats and developing mechanisms to prevent misuse before making a public version available. Fable 5 is the outcome of that effort, offering the same underlying intelligence while limiting access to high-risk capabilities. Fable 5 vs Mythos 5: same foundation, different access rules Anthropic says Claude Fable 5 and Claude Mythos 5 are built on the same underlying model architecture. The key distinction lies in access controls and safety restrictions. Fable 5 has been designed for public use, with safeguards that block or redirect requests in sensitive domains. Mythos 5, meanwhile, remains available only to trusted organisations participating in Project Glasswing and selected security programmes. Anthropic has effectively separated capability from accessibility.While both models possess similar core intelligence, Mythos 5 can operate with fewer restrictions in approved environments, particularly for cybersecurity applications. Fable 5 prioritises safety and broad deployment, whereas Mythos 5 prioritises specialised use cases that require deeper access. This strategy allows Anthropic to commercialise frontier AI while limiting potential misuse.ParameterClaude Fable 5Claude Mythos 5Core architectureMythos-basedSame architecturePublic availabilityYesNoSafety restrictionsExtensiveLimitedCybersecurity functionsRestrictedHigher accessBiology/Chemistry tasksRestrictedBroader capabilityTarget usersDevelopers, enterprises, public usersTrusted organisationsDeploymentClaude and API servicesProject GlasswingRisk profileLowerHigherPurposeGeneral useAdvanced specialised useAccess modelCommercial availabilityControlled access programmeInside Anthropic's safety framework for Fable 5A major part of the Fable 5 launch is Anthropic's emphasis on safety. The company has introduced safeguards that monitor requests related to cybersecurity, biology, chemistry and other potentially sensitive domains. When such requests are detected, Fable 5 can redirect them to Claude Opus 4.8, a less capable but more tightly controlled model. Anthropic says these safeguards are expected to activate in only a small percentage of interactions while allowing most users to benefit from Fable 5's capabilities.The company also conducted extensive testing and evaluations before release. By separating high-risk requests from general-purpose usage, Anthropic hopes to reduce the likelihood of misuse while preserving access to advanced AI capabilities for legitimate users across industries.Why the launch matters for developers and enterprisesThe introduction of Fable 5 could significantly expand access to frontier AI capabilities for developers and businesses. Anthropic positions the model as particularly strong in software engineering, research and complex enterprise workflows. Developers can use it for coding assistance, debugging, automation, document analysis and knowledge-intensive tasks.Because Fable 5 is available through APIs and enterprise offerings, organisations can integrate it into products and internal systems without requiring special approval programmes. The launch also reflects a broader industry trend in which AI companies seek to commercialise increasingly powerful models while implementing layered safeguards. For enterprises, Fable 5 provides access to a model that was previously available only in restricted environments, potentially accelerating innovation across software, finance, healthcare and research sectors. The india angleFor India, the launch of Claude Fable 5 could have implications across the country's large technology ecosystem. Indian software developers, startups, IT services firms and enterprise customers are expected to gain access to a more capable AI platform for coding, automation and research applications. Companies working in software development, business process automation and AI-powered services may benefit from Fable 5's advanced reasoning and engineering capabilities.The model's API availability could also support India's growing generative AI startup ecosystem, which increasingly relies on frontier foundation models to build consumer and enterprise products. At a time when Indian firms are investing heavily in AI adoption, the availability of a more powerful Claude model may create new opportunities for productivity gains and innovation across sectors. Availability, Pricing and Rollout TimelineAnthropic has made Claude Fable 5 available through its Claude platform, APIs and enterprise offerings. According to the company, the model is initially included for Pro, Max, Team and seat-based Enterprise subscribers.However, this arrangement is temporary. Anthropic has indicated that beginning June 23, usage credits may be required to access the model, depending on capacity availability. The company has also announced Claude Mythos 5 alongside Fable 5, though access to the unrestricted version remains limited to approved organisations participating in Project Glasswing and related programmes. Pricing for Fable 5 has been set at $10 per million input tokens and $50 per million output tokens. Broader access expansions are expected to occur gradually as Anthropic evaluates safety and demand. FAQs1. What is Claude Fable 5?Claude Fable 5 is Anthropic's first publicly available Mythos-class AI model, designed for coding, research, reasoning and multimodal tasks.2. Is Claude Fable 5 the same as Claude Mythos 5?Both share the same architecture, but Fable 5 includes additional safety restrictions for public use.3. Why was Mythos previously restricted?Anthropic cited concerns around advanced cybersecurity capabilities and potential misuse.4. Who can use Claude Fable 5?Developers, enterprises and subscribers can access it through Claude services and APIs.5. How much does Claude Fable 5 cost?Anthropic has priced the model at $10 per million input tokens and $50 per million output tokens.end of article