Skip to content

Sieve 2B v2

Sthānika AIDecision modelOpen weights

Sieve 2B v2 by Sthānika AI is a decision model: it answers typed questions with a probability for each allowed answer instead of generating text. It scores 30.8 on the Jev Decision Index (rank 106).

Pricing

No hosted price is listed. The weights are open: the cost is your own compute.

Specifications

Provider
Sthānika AI
Parameters
2B
Released
Oct 2026
Open weights
Yes
Question types
Choice (one of N options), Score (ordered rubric), Yes / no
Mechanism
Decoder LM with a decision readout
Backbone
Qwen/Qwen3.5-2B-Base
Jev-compatible request
No
Runs on CPU
No
License
apache-2.0
Hugging Face
sthanika-ai/Sieve-2B-v2

Benchmarks

Jev Decision IndexECE 0.095
30.8

Jev Decision Index: public 37.3, ECE 0.095, Brier 0.497. Lower ECE means its probabilities can be taken at face value.

Self-reported by the publisher

The model's own claims, often on its own sets or a self-run copy of a public suite. Not independently verified and never ranked alongside measured results.

Jev Decision Index public suite (self-run)Self-reportedDecision Index public balanced skill, edition 0.3 kit (release-v3), self-run
37.3
Decision Index 0.3 public suite (140,178 scored requests; thSelf-reported
37.3
Decision Index 0.3.2 leaderboard (public suite and the maintSelf-reported
30.8

Speed & latency

Median latency per call (Decision Index)
68 ms

Compare Sieve 2B v2 with…

FAQ

How much does Sieve 2B v2 cost?

Sieve 2B v2 has no hosted price listed; its weights are open, so the cost is the compute you run it on.

What questions can Sieve 2B v2 answer?

Sieve 2B v2 answers choice (one of n options), score (ordered rubric), yes / no questions, returning a probability for each allowed answer.

Is Sieve 2B v2 well calibrated?

On the Jev Decision Index its expected calibration error (ECE) is 0.095; below about 0.05, a stated 80% confidence is right roughly 80% of the time.

Is Sieve 2B v2 open weight, and can it run on a CPU?

Yes, Sieve 2B v2 has open weights. It needs a GPU to run at useful speed.

Data as of 2026-10-14. Source: Metatext decision-model catalogue, 2026-10-14 (Jev Decision Index, Hugging Face). Pricing is per 1M tokens (USD); confirm with the provider before production use.