Sthanika’s Sieve 31B Ranks Second Globally on Jev Decision Index

Hyderabad-based Sthanika, the applied AI research lab of the PurpleTalk group, has announced that its Sieve 31B model has secured the second position globally on the Jev Decision Index, an open benchmark hosted on Hugging Face that evaluates AI models designed for rapid decision-making. Sthanika’s decision models ranked ahead of Jev and Perplexity Decider v1.1.
Sieve 31B recorded a Full score of 63.8, combining results from the public test suite with private tests conducted by the benchmark’s maintainers. On the public tests, the model ranked second with a full score of 62.5.
The announcement was made in Hyderabad, Telangana, on October 16, 2026. Sthanika described Sieve 31B as a cost-efficient decision model that can be called by AI agents as one component of a larger system. According to the company, it makes decisions more reliably than general-purpose large language models.
Many AI agents currently send every task to a large language model, including planning and routine decisions that a dedicated decision model can handle at a lower cost. As agents take on more tasks previously performed by people, the associated costs increase. General-purpose LLMs are designed and tuned to converse with people, making them suitable for language-related tasks but expensive when repeatedly choosing between defined options thousands of times a day. By comparison, decision models are extremely inexpensive for such tasks.
The wider Sieve family also includes Sieve 2B Plus, Sieve 4B and Sieve 9B Plus, each of which placed among the top models in its respective size class on the same leaderboard. Developers can run smaller models on their own on-premise GPUs for everyday tasks requiring privacy and the processing of large volumes of work. They can then switch to larger, more powerful models when tasks demand more advanced capabilities.
While most discussions about AI focus on models that write, summarise and converse, much of the AI operating inside real-world software performs a less visible function: making small decisions thousands of times a day. Sieve has been developed specifically for this decision-making role.
The model reads a question and the available choices as presented, then returns a defined answer accompanied by a confidence score that software can inspect and act upon. This approach is suited to applications where consistency, speed and the ability to audit each decision are as important as accuracy.
"Sieve 31B shows that a focused model fine-tuned in India can perform among the best in the world at this work. The private tests matter most to us. They come from domains outside the public suite, and Sieve generalised better than most of the field. It learned the task, not the test," said Sridhar Muppidi, Co-founder of PurpleTalk.
Sieve builds on Sthanika’s earlier work on compact, confidence-aware decision models, which are already deployed inside PurpleTalk products, including Ello and Mina. Within these products, the models handle routine judgements faster and more economically than general-purpose generative models.
The research lab also expects decision models to enable a new category of agentic systems offering greater reliability, more granular control and low latency in areas such as voice, customer experience and robotics.
"We believe decision models will open up a new category of agentic systems: more reliable, more controllable and fast enough for voice, customer experience and robotics. We are publishing everything needed to reproduce our results, and we invite the research community to test it, challenge it and build on it," said Bharat Lingam, CEO & Co-founder, PurpleTalk.
The models and evaluation artifacts have been made available through a link provided in the announcement, while the complete results are available on the Jev Decision Index.
Sthanika is a Hyderabad-based AI research lab that develops open-weight models for problems general-purpose models were not designed to solve. Its name is derived from the Sanskrit word for "local." The lab operates as part of the PurpleTalk group, and further information is available at sthanika.ai.
Founded in 2008, PurpleTalk builds AI companies and services. The group has worked across healthcare, agriculture, financial services and retail. Its portfolio includes (x) cube LABS, Sthanika, Ello and Mina.
Sthanika’s second-place global ranking for Sieve 31B highlights its focus on specialised AI decision-making models, with the company emphasising cost efficiency, confidence-aware outputs and reproducible evaluation results as it invites the research community to independently test and build on its work.