Anthropic Launches Claude Fable 5.1 With Lower Cache Costs
Anthropic launches Claude Fable 5.1 with lower cache costs, improved efficiency, and advanced AI capabilities designed for faster.
Anthropic released Claude Fable 5.1 and Claude Mythos 5.1 on Sept. 1, introducing lower cache costs, fewer false-positive safety interventions, and new enterprise privacy controls as the company competes for broader AI adoption across business workloads.
Sept. 2, 2026: Anthropic released Claude Fable 5.1 and Mythos 5.1 on Sept. 1. Fable 5.1 is available through Anthropic’s API, Claude plans, and supported cloud platforms, while Mythos 5.1 remains limited to vetted organizations working in cybersecurity and life sciences, according to Anthropic’s Fable 5.1 and Mythos 5.1 announcement.
Key Stats
- Anthropic Claude Fable 5.1 retains standard pricing of $10 per million input tokens and $50 per million output tokens.
- Cache-read pricing has fallen 75% to $0.25 per million tokens.
- Anthropic estimates the change could reduce typical workload costs by around 25%.
- Anthropic estimates savings of up to approximately 45% for highly agentic workloads.
- Anthropic said updated cybersecurity safeguards generate around 60% fewer interventions.
- Anthropic said biology safeguards intervene 85% less often on benign elementary biology and medical requests.
The pricing change applies to cached context rather than Fable 5.1’s standard input and output token rates. Anthropic’s cost and safeguard figures are company estimates and have not been independently benchmarked. Lower cache-read costs can reduce spending for AI agents and enterprise workflows that repeatedly reuse contextual information, according to Anthropic’s Claude Fable model page.
The reduction comes as enterprise AI costs face growing scrutiny, with companies under greater pressure to control AI spending and prove measurable returns.
Anthropic has also adjusted Fable 5.1’s safeguards to reduce false-positive restrictions on legitimate requests. TechCrunch reported that the new release combines lower workload costs with fewer safeguard interventions while retaining controls around higher-risk cybersecurity and biological activity.
Alongside the model release, Anthropic announced Enterprise Frontier Safeguards (EFS), a system intended to give organizations greater control over how enterprise activity data and misuse monitoring are handled.
Under the planned system, eligible customers will be able to keep activity data within cloud infrastructure they control, while human review can be handled by the customer rather than Anthropic. EFS is expected to roll out in phases later in 2026. Until then, eligible enterprise customers can use Fable 5.1 with zero data retention, according to Anthropic’s Enterprise Frontier Safeguards announcement.
The release reflects a shift in frontier AI competition beyond benchmark performance. Enterprise buyers increasingly have to compare model capability with operating cost, privacy controls, safeguards, and deployment requirements.
For B2B and technology leaders, Anthropic Claude Fable 5.1 shows how AI procurement is beginning to resemble traditional enterprise software buying. As AI platform and model spending continues to grow, performance increasingly sits alongside cost, governance, and practical deployment requirements in enterprise buying decisions.
