Claude Fable 5: The first Mythos model is powerful, expensive, and heavily filtered

The 9% request block rate is the number that actually matters for enterprise adoption decisions, and it's conspicuously absent from most coverage framing. A nearly one-in-ten rejection rate at the API layer isn't a footnote, it's a material operational constraint that changes how builders architect around the model.

The premium pricing and mandatory data retention policies Anthropic is imposing here connect directly to the infrastructure cost pressures we covered the same day in the China underwater data center story. As frontier inference costs rise, the question of where compute runs and who controls the data trail becomes a single bundled decision for enterprise buyers, not two separate ones. Anthropic is effectively raising the floor on what it costs to build at the capability frontier, which accelerates the incentive for large operators to seek alternative infrastructure arrangements rather than absorb both the token costs and the policy overhead.

Watch whether any of the major cloud providers (AWS, Google Cloud, Azure) respond within the next 60 days by announcing preferential pricing or relaxed retention terms for their own frontier model tiers. If they do, Anthropic's Mythos positioning will have functioned as a competitive opening rather than a durable moat.