Anthropic says Fable 5.2 nearly doubles its predecessor’s hardware-design score. Token prices fell 20%, but the cost per task doubled at full effort.

Anthropic reported that its Fable 5.2 model scored 83.5 on BenchCAD 2.1, up from 44.4 for Fable 5.1, nearly doubling its predecessor's hardware-design benchmark result. Token prices fell 20%, but the cost per index task at maximum effort rose to $26.54 from $12.99, meaning the cost per task more than doubled at full effort.

Fable 5.2 scored 83.5 on BenchCAD 2.1, up from Fable 5.1’s 44.4. Token prices fell 20%, but cost per index task at max effort rose to $26.54 from $12.99. The post Anthropic says Fable 5.2 nearly doubles its predecessor’s hardware-design score. Token prices fell 20%, but the cost per task doubled at full effort. https://www.rdworldonline.com/anthropic-says-fable-5-2-nearly-doubles-its-predecessors-hardware-design-score-token-prices-fell-20-but-the-cost-per-task-doubled-at-full-effort/ appeared first on Research & Development World https://www.rdworldonline.com .