Samaya AI Launches Sterling Finance Research Model
The new model outperforms existing industry benchmarks while significantly reducing operational costs.
Updated on Oct. 30, 2026 in Artificial Intelligence
Live Poll
Do you believe smaller, specialized AI models can reliably outperform large, general-purpose systems?
Samaya AI has released the Samaya Sterling finance research agent, a model designed to handle complex financial queries. It demonstrates superior performance on the FrontierFinance benchmark compared to the Claude Fable 5.1 model.
Why it matters
By delivering higher accuracy at a fraction of the cost of current industry leaders, this model aims to improve the efficiency of automated financial analysis. It represents a significant step in optimizing specialized research tools for enterprise use.
Samaya Sterling utilizes a 30 billion parameter mixture-of-experts architecture with 3 billion active parameters. The model calls 37 distinct tools and costs $0.61 per question, compared to $6.89 for Claude Fable 5.1.
The players
Samaya AI
The artificial intelligence research company responsible for the development and release of the Samaya Sterling financial agent.
Thinking Machines Lab
A research entity providing the Tinker training API used to conduct the model development process.
NVIDIA
A prominent technology company whose Nemotron 3.5 Lightning model serves as the foundational architecture for the new agent.
The details
The model is a LoRA post-trained version of NVIDIA's Nemotron 3.5 Lightning, developed using the Tinker training API from Thinking Machines Lab. It employs a system of helper models, a condenser, and a compactor to manage context and summarize traces for complex financial tasks.
Timeline
October 30, 2026: Samaya AI released the Samaya Sterling model.
The Tech Race
This release follows the trend of creating specialized, cost-effective agents built atop foundational models like NVIDIA's Nemotron 3.5 Lightning. It signals a move away from general-purpose systems toward highly optimized, domain-specific AI that performs better at a lower price point.
Financial researchers and analysts may see faster turnaround times and lower costs for automated reporting and data analysis. This shift allows for the integration of more sophisticated research tools into existing financial workflows at a significantly reduced price.
The takeaway
Specialized finance agents are becoming increasingly efficient by leveraging mixture-of-experts architectures. Developers should monitor how these tools integrate with existing financial data platforms to optimize their own research capabilities.
Further reading
For more on the current state of specialized AI models, see Artificial Intelligence.
Source note: This article includes information reported by MarkTechPost.
Live Poll
Do you believe smaller, specialized AI models can reliably outperform large, general-purpose systems?
