Hopper 12B (trained) vs Celeris-1 Decision
Hopper 12B (trained) and Celeris-1 Decision have no measured difference in accuracy, calibration or cost in our data yet.
| Spec | Hopper 12B (trained) | Celeris-1 Decision |
|---|---|---|
| Provider | HopitAI | Celeris |
| Decision Index score | 47.3 | — |
| Calibration error (ECE) | 0.076 | — |
| Median latency (p50) | 290 ms | — |
| Cost per 1,000 decisions | — | $0.0080 |
| Parameters | 12B | — |
| Open weights | Yes | No |
| Question types | Choice (one of N options), Score (ordered rubric), Yes / no | Choice (one of N options), Score (ordered rubric), Yes / no |
| Mechanism | Decoder LM with a decision readout | Diffusion LM |
| License | research-and-demo | proprietary |
FAQ
Is Hopper 12B (trained) or Celeris-1 Decision more accurate?
At least one of the two has no Jev Decision Index score yet, so their accuracy cannot be compared here.
Which is better calibrated, Hopper 12B (trained) or Celeris-1 Decision?
No calibration difference between Hopper 12B (trained) and Celeris-1 Decision is recorded yet.
Is Hopper 12B (trained) or Celeris-1 Decision 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.