VERIFIEDNov 222026
Is Raw Qwen3 1.7B direct logits state of the art?#444Best current rank: #444 in Decision models, as of Nov 22, 2026.
As of Nov 22, 2026, Raw Qwen3 1.7B direct logits is ranked #444 in Decision models (since Nov 22, 2026; sources JevBench; methodology v1.0). Ranked #444 in Decision models.
At a glance
Facts and scores
Raw Qwen3 1.7B direct logits is an open-weights (Apache-2.0) text model from Alibaba / Qwen; 1.7B parameters and no public price.
How it scores
- On JevBench price it scores 1.1%, #29 of 134 (top 22%); 1.0 pts behind verdict-small (Manavarya09, multilingual-e5-small 118M); level with WaterSheep (Samrat Dutta, ModernBERT-base encoder with calibrated heads, served by the author's own `watersheep --serve`) above and Decision 2B (FlyMy.AI, v59) is 0.2 pts below.
- On JevBench it scores 8.3 (CI 7.8–8.8), #132 of 134 (top 99%); 74.5 behind Perplexity Decider v1.1 27B (Qwen3.8-27B, noncausal decision readout); ClassOne Gemma 4 E2B (Gemma-4-E2B backbone with ClassOne decision heads, schema-constrained single forward pass, self-hosted) is 6.9 above and Raw Qwen3 0.6B direct logits is 0.1 below.
History
First result on isitsota: Oct 8, 2026. Best overall rank so far: #444 in Decision models, first held Nov 22, 2026.
$0.0112 per 1,000 decisions.
Receipts
Evidence
| Source | Score | Date | Type | Receipt |
|---|---|---|---|---|
| JevBench Raw Qwen3 1.7B direct logits | 8.3 7.8–8.8 | 2026-10-08 | INDEP. | snap_jevbench_2026-10-08_v1.6.1|jevbench-v1.6.1 |
Reign history
Where it sits in the succession
Crown intervals for this model are still empty.
Honest about the gaps
Where it is not the crown
| Category | This model | Rank | Leader |
|---|---|---|---|
| Decision models | #444 in this category | #444 | karpo-decision-31B |
Links
Related pages
Is Raw Qwen3 1.7B direct logits SOTA?
Best current rank: #444 in Decision models, as of Nov 22, 2026.
What is Raw Qwen3 1.7B direct logits best at?
As of Nov 22, 2026, Raw Qwen3 1.7B direct logits is ranked #444 in Decision models (since Nov 22, 2026; sources JevBench).