# Is Sieve 27B v2 SOTA?

Best current rank: #104 in Decision models, as of Nov 22, 2026.

As of Nov 22, 2026, Sieve 27B v2 is ranked #104 in Decision models (since Nov 22, 2026; sources Decision Index Full; methodology v1.0). Ranked #104 in Decision models.

Canonical: https://isitsota.com/m/sieve-27b-v2
Last updated: 2026-11-21
Methodology v1.0. Computed automatically.

- Decision models: #104 in Decision models — Ranked #104 in Decision models.

## At a glance

Sieve 27B v2 is an open-weights text model from Sthānika AI; 28B parameters and no public price.

### How it scores

- On Decision Index Brier it scores 28.3%, #52 of 419 (top 13%); 7.9 pts behind lg-decision-4B; GLiDE 9B is 0.1 pts above and level with Sloth-27B v1 below.
- On Decision Index Full it scores 59.9 (CI 59.4–60.4), #108 of 420 (top 26%); 14.0 behind karpo-decision-31B; level with OneJev1.5-27B above and mogan-decision-31B-it is 0.1 below.
- On Decision Index ECE it scores 5.3%, #162 of 420 (top 39%); 4.2 pts behind KnowLine 9B Gen2; vLLM-SR d3-turbo v1.1 (27B) is 0.1 pts above and liquamen (Gemma 4 31B) is 0.1 pts below.

### History

First result on isitsota: Nov 21, 2026. Best overall rank so far: #104 in Decision models, first held Nov 22, 2026.


## Related

- Lab: [Sthānika AI](https://isitsota.com/labs/sth-nika-ai.md)
- Category: [Decision models](https://isitsota.com/c/decision.md)
- JSON: https://api.isitsota.com/v1/models/sieve-27b-v2
