Users Dispute 10T Model Claims for Fable

@scaling01 posted that Mythos Preview exceeds Mythos 5 and Fable in size, citing its higher price per output token and stronger benchmark results. @teortaxesTex replied that the evidence is weak, noting the lack of user reports describing Fable as a distilled version and suggesting the larger size claim may be disinformation. The thread also covered relative rankings of AI labs and questions about inference costs.