Anthropic prices both Claude Fable 5 and Mythos 5 at $10 per 1M input tokens and $50 per 1M output tokens, less than half the price of Claude Mythos Preview
ZDNET's key takeaways — Anthropic is releasing Claude Fable 5 for general users. — Fable 5 uses Mythos-class power with safety controls.
Context & Ripple Effects
Anthropic is separating access to its newest Mythos-class capability: Fable 5 is being released broadly with safety controls, while Mythos 5 is directed to trusted organizations. The company says Fable’s classifiers can route sensitive sessions to Claude Opus 4.8, with such fallbacks occurring in fewer than 5% of sessions.
The shared price for Fable 5 and Mythos 5 is a sharp reduction from Claude Mythos Preview. That makes pricing part of Anthropic’s broader attempt to widen deployment while retaining differentiated safety and access controls.
First-order effects
- Developers and organizations using Fable 5 or Mythos 5 face materially lower per-token costs than under the Mythos Preview pricing, reducing the cost of deploying these models in token-intensive workloads.
- Anthropic can present Fable 5 as a broadly available, lower-cost route to Mythos-class performance while reserving Mythos 5 access for trusted organizations.
Second-order effects
- Lower pricing raises pressure on rival model providers to compete more directly on the cost of high-capability inference, rather than treating premium capability as a separately priced tier.
- Customers may reassess whether restricted access to a top model is necessary when a public alternative offers comparable positioning with classifier-based controls and fallback behavior.
Third-order effects
- If high-capability models continue to be sold at lower inference prices, differentiation is likely to shift toward reliability, safety-routing design, data-handling terms, and deployment access rather than headline model capability alone.
- Anthropic’s split between broad controlled access and trusted-organization access points toward a market where safety controls become part of product segmentation; adoption may depend as much on enterprise policy compatibility as on price.
The trend: Frontier-model vendors are pushing powerful models into wider use at lower unit costs while using access tiers and safety systems to manage the resulting risk.