GLiNER2.5-Decide vs tasksource Decider nano v4 (150M)
tasksource Decider nano v4 (150M) is more accurate on the Jev Decision Index, tasksource Decider nano v4 (150M) is better calibrated. Choose on whichever of accuracy, calibration and cost matters most for your decisions.
| Spec | GLiNER2.5-Decide | tasksource Decider nano v4 (150M) |
|---|---|---|
| Provider | Fastino | tasksource |
| Decision Index score | 10.1 | 19.9 |
| Calibration error (ECE) | 0.088 | 0.070 |
| Median latency (p50) | 23 ms | 13 ms |
| Cost per 1,000 decisions | — | — |
| Parameters | 486M | 149M |
| Open weights | Yes | Yes |
| Question types | Choice (one of N options), Score (ordered rubric), Yes / no, Extraction (typed spans) | Choice (one of N options), Score (ordered rubric), Yes / no |
| Mechanism | Span encoder (GLiNER family) | Cross-encoder (NLI) |
| License | apache-2.0 | apache-2.0 |
FAQ
Is GLiNER2.5-Decide or tasksource Decider nano v4 (150M) more accurate?
tasksource Decider nano v4 (150M) scores higher on the Jev Decision Index (19.9 vs 10.1).
Which is better calibrated, GLiNER2.5-Decide or tasksource Decider nano v4 (150M)?
tasksource Decider nano v4 (150M) has the lower expected calibration error (0.070 vs 0.088), so its stated confidence is closer to how often it is right.
Is GLiNER2.5-Decide or tasksource Decider nano v4 (150M) cheaper per 1,000 decisions?
A hosted price is not recorded for both models; open-weight models cost the compute you run them on.
Cost per 1,000 decisions assumes 200 input tokens per decision. Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.