Samaya introduces Sterling, claiming a higher FrontierFinance score than Claude Fable 5.1 at 11× lower cost

Samaya says Sterling is a post-trained 30B mixture-of-experts model with 3B active parameters. On its FrontierFinance financial-research benchmark, Samaya reports a score of 58.9%, compared with 55.9% for Claude Fable 5.1, at $0.61 versus $6.89 per question. Samaya says two supervised fine-tuning stages and two reinforcement-learning stages raised the base model’s score from 42.5%.