ENPT

Cerno (27B / 9B / 4B)

Last updated:

Open52–52ms$0/M input

Quadrant scores

See the full quadrant

Scored with the public quadrant rubric (maturity × capability; bubble = adoption). Revise as evidence lands.

  • Maturity5.2/10

    Apache-2.0 weights in 27B, 9B and 4B (Qwen bases, FP8 for the 27B) with one shared inference code path and published Decision Index runs. Company-run, no hosted API or SLA; few integrations.

  • Capability7.6/10

    Decision Index 0.4.1 (22 Oct) Full 61.68 (#31) for the 27B (9B and 4B not listed on 0.4.1), above Jev 1.13 (59.22). Choice (up to 255), yes/no and score in one pass. Author board-style ECE 5.7 pts and 52 ms median on one H200.

  • Adoption1/100

    About 180 downloads and 9 likes across the three repos (20 Oct). No integrations or production users found.

Vendor claims

publicado (OK do Fabricio 20/10 BRT). Scores are calculated from scratch.

What it is

Apertix AI’s open Jev-style decision family (27B, 9B, 4B, Apache-2.0) with calibrated probabilities in one pass. Decision Index 0.4 Full 62.64 (#26) for the 27B. Source: huggingface.co/apertixai/cerno-27b. The card’s “second of 145” refers to the 0.3 public suite run by the authors, not to the 0.4 Full ranking.

Board result

Board Row Result
Decision Index 0.4 (20 Oct 2026) Cerno (27B / 9B / 4B) Full 62.64 (#26); Jev 1.13 61.26 (#34), Jev 1.14 62.22 (#28)

Not checked against JevBench or Artificial Analysis for this page; the boards are never combined.

Vendor claims (UnverifiedClaim)

  • 27B: second of 145 entrants on the Decision Index 0.3 public suite on a like-for-like basis; board-style ECE 5.7 pts; golden set 92.7% accuracy, ECE 0.022 (source)
  • 52 ms median per request on one H200, one request at a time (source)

Scores (from scratch, 20 Oct 2026)

  • Capability 7.6 = types × 0.30 + calibration × 0.25 + latency × 0.25 + accuracy × 0.20 = 10 × 0.30 + 6 × 0.25 + 6 × 0.25 + 8 × 0.20 = 7.60. Accuracy: 8 × 62.64 ÷ 61.26 = 8.18 → 8.
  • Maturity 5.2 = availability × 0.30 + docs × 0.25 + integrations × 0.25 + support × 0.20 = 7 × 0.30 + 7 × 0.25 + 3 × 0.25 + 3 × 0.20 = 5.20.
  • Adoption 1 = engagement × 0.35 + ecosystem × 0.35 + production × 0.30 = 2 × 0.35 + 1 × 0.35 + 0 × 0.30 = 1.05 → 1.

Board audit (22 Oct 2026, DI 0.4.1)

publicado (OK do Fabricio 22/10 BRT).

Decision Index 0.4.1 (22 Oct; reference Jev 1.13 Full 59.22, was 61.26 on 0.4) lists this page’s row at Full 61.68 (#31). Accuracy rule: 8 × 61.68 ÷ 59.22 = 8.33 → 8 (was 8). Capability unchanged. Maturity and adoption unchanged.