Kev 0.8B r15

Jared PalmerDecision modelOpen weights

Kev 0.8B r15 by Jared Palmer is a decision model: it answers typed questions with a probability for each allowed answer instead of generating text. It scores 13.7 on the Jev Decision Index (rank 133).

Pricing

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

Specifications

Provider: Jared Palmer
Parameters: 870M
Released: Sep 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/Qwen3.5-0.8B-Base
Jev-compatible request: Yes
Runs on CPU: Yes
License: apache-2.0
Hugging Face: jaredpalmer/kev-0.8b

Benchmarks

Jev Decision IndexECE 0.040: 13.7
Jev Decision Indexedition 0.2.1 (earlier edition): 14.6
PT-BR typed decisions: 50.7

Jev Decision Index: public 16.8, ECE 0.040, Brier 0.604. 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.

breadth-v1 test: 14 held-out public datasets (3,089 questionSelf-reported: 0.586
documents-v1 test (936 questions on CFPB complaint narrativeSelf-reported: 0.851
hard-v1 test (1,088 questions; programmatic labels, held-outSelf-reported: 0.665
tasksource-heldout-v1 development: 17 of 24 held-out task faSelf-reported: 0.515
transfer-v4 test: six never-trained public sources and held-Self-reported: 0.697

Speed & latency

Median latency per call (Decision Index)

45 ms

Compare Kev 0.8B r15 with…

Kev 0.8B r15 vs amos-1-large—/1M in Kev 0.8B r15 vs amos-1-medium—/1M in Kev 0.8B r15 vs Celeris-1 Decision$0.04/1M in Kev 0.8B r15 vs ScaleDown Classify-1$0.04/1M in Kev 0.8B r15 vs Cloudflare clef—/1M in Kev 0.8B r15 vs Cloudflare clef-flash—/1M in

FAQ

How much does Kev 0.8B r15 cost?

Kev 0.8B r15 has no hosted price listed; its weights are open, so the cost is the compute you run it on.

What questions can Kev 0.8B r15 answer?

Kev 0.8B r15 answers choice (one of n options), score (ordered rubric), yes / no questions, returning a probability for each allowed answer.

Is Kev 0.8B r15 well calibrated?

On the Jev Decision Index its expected calibration error (ECE) is 0.040; below about 0.05, a stated 80% confidence is right roughly 80% of the time.

Is Kev 0.8B r15 open weight, and can it run on a CPU?

Yes, Kev 0.8B r15 has open weights. It is small enough to run on a CPU.

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