ENPT

AJev (Gemma 4 31B LoRA lm6q)

Last updated:

Open139–540ms$0/M input

Quadrant scores

See the full quadrant

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

  • Maturity4.8/10

    Apache-2.0 LoRA (r 64) for google/gemma-4-31B-it with an inference package (cmzy/ajev-infer, vLLM) and a local /v1/systemone server. Needs about 62 GB of GPU memory. Independent author, no hosted API, no SLA.

  • Capability7.1/10

    Decision Index 0.4.1 (22 Oct) Full 67.57 (#7), above Jev 1.13 (59.22). Yes/no, choice (up to 255 options) and score, one forward pass per question. No ECE published. Author latency on one RTX PRO 6000: median 139 ms, p80 540 ms.

  • Adoption1/100

    About 19 downloads and no likes on the repo (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

Independent Jev-style LoRA for Gemma 4 31B: yes/no, choice and score with a probability per option. Decision Index 0.4 Full 69.05 (#5). Source: huggingface.co/andyzhang232/ajev-gemma4-31b-lm6q. The same author has other AJev adapters on the board: lora1 (Gemma 4 26B-A4B) 55.84 and lora7 (Gemma 4 12B) 49.11. This page is scored on lm6q.

Board result

Board Row Result
Decision Index 0.4 (20 Oct 2026) AJev (Gemma 4 31B LoRA lm6q) Full 69.05 (#5); 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)

  • Latency one request at a time on one RTX PRO 6000 (vLLM 0.30.0, bf16, 506-request sample): median 139 ms, mean 620 ms, p80 540 ms (source)

Scores (from scratch, 20 Oct 2026)

  • Capability 7.1 = types × 0.30 + calibration × 0.25 + latency × 0.25 + accuracy × 0.20 = 10 × 0.30 + 4 × 0.25 + 6 × 0.25 + 8 × 0.20 = 7.10. Accuracy: 8 × 69.05 ÷ 61.26 = 9.02, capped at 8.
  • Maturity 4.8 = availability × 0.30 + docs × 0.25 + integrations × 0.25 + support × 0.20 = 7 × 0.30 + 6 × 0.25 + 3 × 0.25 + 2 × 0.20 = 4.75 → 4.8.
  • Adoption 1 = engagement × 0.35 + ecosystem × 0.35 + production × 0.30 = 1 × 0.35 + 1 × 0.35 + 0 × 0.30 = 0.70 → 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 67.57 (#7). Accuracy rule: 8 × 67.57 ÷ 59.22 = 9.13 → 8 (was 8). Capability unchanged. Maturity and adoption unchanged.