Developer and publisher of Claude Fable 5.1 / Mythos 5.1; set pricing, safeguard policy, and the zero-data-retention offering, following the Fable 5 'nerfed'/'caged' backlash two months earlier.
ZV
Zvi Mowshowitz (Don't Worry About the Vase)
Independent AI analyst who published separate capabilities and system-card writeups on Fable 5.1/Mythos 5.1, calling the model excellent and well-rounded while flagging modest benchmark gains, a narrow early Astra comparison, and safety regressions that lead him to doubt Anthropic's own Mythos 5.1 safety-tier claim.
AM
Amazon Web Services (AWS)
Cloud partner that made Fable 5.1 available on its platform at launch; AWS security researchers were also the ones who discovered the Fable 5 jailbreak that triggered June's export-control suspension.
Independent benchmarking firm that found Fable 5.1 costs 20% more per task than Fable 5 at maximum effort ($3.76 vs $3.14) because it produces 1.7x more output tokens, complicating the headline efficiency story.
Competitor whose GPT-6 Astra model is the most frequent head-to-head comparison point for Fable 5.1 in third-party benchmarks, with results split depending on the test.