Cloudflare clef vs Nace.AI Drex v2.0 (35B-A3B)
Nace.AI Drex v2.0 (35B-A3B) is more accurate on the Jev Decision Index, Nace.AI Drex v2.0 (35B-A3B) is better calibrated. Choose on whichever of accuracy, calibration and cost matters most for your decisions.
| Spec | Cloudflare clef | Nace.AI Drex v2.0 (35B-A3B) |
|---|---|---|
| Provider | Cloudflare | Nace.AI |
| Decision Index score | 53.1 | 55.9 |
| Calibration error (ECE) | 0.040 | 0.018 |
| Median latency (p50) | 102 ms | 276 ms |
| Cost per 1,000 decisions | — | $0.012 |
| Parameters | 28B | 35B |
| Open weights | Yes | Yes |
| 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 | Decoder LM with a decision readout |
| License | apache-2.0 | nace-ai-open-rail-m |
FAQ
Is Cloudflare clef or Nace.AI Drex v2.0 (35B-A3B) more accurate?
Nace.AI Drex v2.0 (35B-A3B) scores higher on the Jev Decision Index (55.9 vs 53.1).
Which is better calibrated, Cloudflare clef or Nace.AI Drex v2.0 (35B-A3B)?
Nace.AI Drex v2.0 (35B-A3B) has the lower expected calibration error (0.018 vs 0.040), so its stated confidence is closer to how often it is right.
Is Cloudflare clef or Nace.AI Drex v2.0 (35B-A3B) 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.