Benchmarks
More
OverviewBenchmaxxingCompareChartsCost vs CapabilityProviders per ModelProvider explorerGatewaysEU & SovereignJev-class modelsImage JevBenchAbout
JevBench by Benchmark Heaven · v1.4.2.2 · individual system

GPT-6 Luna (default medium reasoning effort)

Instruction model, JSON schema · by OpenAI · API endpoint saw sealed item text

JevBench v1.4.2.2 score

33.3

Rank #37 of 91 ranked systems.

30.0 points behind Jev 1.13.0's 63.3.

Where it sits among the 91 ranked systems. The marked tick is Jev 1.13.0 (63.3).

Published axes

intelligence
97.4
calibration
93.5
speed
72.6
cost
36.0

Against Jev 1.13.0

Four radars compare this fixed pair across the score axes, accuracy per tier, and accuracy by family on the hard tier and sealed set. Further out is better on every spoke.

  • A: GPT-6 Luna — Instruction model, JSON schema · Score 33.3 (#37)
  • B: Jev 1.13.0 — Jev (TypeSafe, closed) · Score 63.3 (#4)

The four score axes

Radar: the four score axes, two systemsThe four score axes, GPT-6 Luna vs Jev 1.13.0. Intelligence: 97.4 vs 53.1; Calibration: 93.5 vs 76.3; Speed: 72.6 vs 83.3; Cost: 36.0 vs 52.0.50100Intelligence97.4 · 53.1Calibration93.5 · 76.3Speed72.6 · 83.3Cost36.0 · 52.0
0–100, the values in the table. A label-only system has no calibration (counted as 0).

Accuracy per tier, incl. sealed

Radar: accuracy per tier, incl. sealed, two systemsAccuracy per tier, incl. sealed, GPT-6 Luna vs Jev 1.13.0. Easy: 100% vs 100%; Standard: 100% vs 99%; Judge: 97% vs 95%; Hard: 99% vs 74%; Sealed: 95% vs 37%.50100Easy100% · 100%Standard100% · 99%Judge97% · 95%Hard99% · 74%Sealed95% · 37%
Share correct per tier; Sealed = the 308 private decisions, aggregate only.

Current question set by family (hard + sealed)

Radar: current question set by family (hard + sealed), two systemsCurrent question set by family (hard + sealed), GPT-6 Luna vs Jev 1.13.0. Ambiguous / abstain: 96% vs 43%; Judge: 91% vs 54%; Long policy: 97% vs 44%; Multi-hop: 99% vs 64%; Probability: 100% vs 63%; Temporal / numeric: 97% vs 28%; Trade-off: 97% vs 55%; Routing: 100% vs 100%; Trap / adversarial: 100% vs 83%; Paraphrase: 100% vs 64%; Safety judge: 94% vs 38%.50100Ambiguous /abstain96% · 43%Judge91% · 54%Long policy97% · 44%Multi-hop99% · 64%Probability100% · 63%Temporal /numeric97% · 28%Trade-off97% · 55%Routing100% · 100%Trap /adversarial100% · 83%Paraphrase100% · 64%Safety judge94% · 38%
Share correct per family across the 220 hard-tier decisions (public and held out) and the 308 sealed decisions of v1.4, pooled; Routing is hard-tier only, Paraphrase and Safety judge sealed only.

Sealed set by family

Radar: sealed set by family, two systemsSealed set by family, GPT-6 Luna vs Jev 1.13.0. Ambiguous / abstain: 95% vs 30%; Judge: 90% vs 34%; Long policy: 95% vs 28%; Multi-hop: 97% vs 45%; Paraphrase: 100% vs 64%; Probability: 100% vs 50%; Safety judge: 94% vs 38%; Temporal / numeric: 95% vs 29%; Trade-off: 96% vs 38%; Trap / adversarial: 100% vs 42%.50100Ambiguous /abstain95% · 30%Judge90% · 34%Long policy95% · 28%Multi-hop97% · 45%Paraphrase100% · 64%Probability100% · 50%Safety judge94% · 38%Temporal /numeric95% · 29%Trade-off96% · 38%Trap /adversarial100% · 42%
Share correct within each sealed family — system-level aggregates; the items stay private.

Availability and evidence

Openness
Unknown in the published row
License note
proprietary API
Cost evidence
measured; the board’s row disclosure contains the published basis.
Endpoint condition
API measurement: sealed item text (no golds) was sent to OpenAI
Note on this row
OpenAI direct API baseline; reasoning effort default medium; strict JSON-schema probability response; temperature unset; max_completion_tokens=4096; price cost from returned usage at official standard list rates.

From the public v1.4.2.2 aggregate. Scores and ranks can change when a new release is published.

Read the full board and published method. The overall score is a composite, not raw accuracy.