GEV-26B-Decide
GEV-26B-Decide by AutoTrust AI Lab is a decision model: it answers typed questions with a probability for each allowed answer instead of generating text. It scores 56.2 on the Jev Decision Index (rank 27).
Pricing
No hosted price is listed. The weights are open: the cost is your own compute.
Specifications
- Provider
- AutoTrust AI Lab
- Parameters
- 26B
- 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
- google/gemma-4-26B-A4B-it
- Jev-compatible request
- Yes
- Runs on CPU
- No
- License
- apache-2.0
- Hugging Face
- autotrust/GEV-26B-Decide
Benchmarks
- Jev Decision IndexECE 0.072
- 56.2
Jev Decision Index: public 58.0, ECE 0.072, Brier 0.353. 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.
- Decision Index suite 0.2 (edition 0.2.1)Self-reported
- 62.5
Speed & latency
Compare GEV-26B-Decide with…
FAQ
How much does GEV-26B-Decide cost?
GEV-26B-Decide has no hosted price listed; its weights are open, so the cost is the compute you run it on.
What questions can GEV-26B-Decide answer?
GEV-26B-Decide answers choice (one of n options), score (ordered rubric), yes / no questions, returning a probability for each allowed answer.
Is GEV-26B-Decide well calibrated?
On the Jev Decision Index its expected calibration error (ECE) is 0.072; below about 0.05, a stated 80% confidence is right roughly 80% of the time.
Is GEV-26B-Decide open weight, and can it run on a CPU?
Yes, GEV-26B-Decide 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, Benchmark Heaven JevBench, Hugging Face). Pricing is per 1M tokens (USD); confirm with the provider before production use.