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

OpenJev (thinking, BF16)

Jev rebuild · by razorback16

JevBench v1.4.2.2 score

14.8

Rank #70 of 91 ranked systems.

48.5 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
58.1
calibration
58.1
speed
76.1
cost
27.8

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: OpenJev — Jev rebuild · Score 14.8 (#70)
  • 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, OpenJev vs Jev 1.13.0. Intelligence: 58.1 vs 53.1; Calibration: 58.1 vs 76.3; Speed: 76.1 vs 83.3; Cost: 27.8 vs 52.0.50100Intelligence58.1 · 53.1Calibration58.1 · 76.3Speed76.1 · 83.3Cost27.8 · 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, OpenJev vs Jev 1.13.0. Easy: 100% vs 100%; Standard: 100% vs 99%; Judge: 95% vs 95%; Hard: 78% vs 74%; Sealed: 42% vs 37%.50100Easy100% · 100%Standard100% · 99%Judge95% · 95%Hard78% · 74%Sealed42% · 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), OpenJev vs Jev 1.13.0. Ambiguous / abstain: 55% vs 43%; Judge: 80% vs 54%; Long policy: 37% vs 44%; Multi-hop: 60% vs 64%; Probability: 52% vs 63%; Temporal / numeric: 36% vs 28%; Trade-off: 55% vs 55%; Routing: 100% vs 100%; Trap / adversarial: 90% vs 83%; Paraphrase: 57% vs 64%; Safety judge: 69% vs 38%.50100Ambiguous /abstain55% · 43%Judge80% · 54%Long policy37% · 44%Multi-hop60% · 64%Probability52% · 63%Temporal /numeric36% · 28%Trade-off55% · 55%Routing100% · 100%Trap /adversarial90% · 83%Paraphrase57% · 64%Safety judge69% · 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, OpenJev vs Jev 1.13.0. Ambiguous / abstain: 38% vs 30%; Judge: 68% vs 34%; Long policy: 25% vs 28%; Multi-hop: 39% vs 45%; Paraphrase: 57% vs 64%; Probability: 36% vs 50%; Safety judge: 69% vs 38%; Temporal / numeric: 30% vs 29%; Trade-off: 35% vs 38%; Trap / adversarial: 67% vs 42%.50100Ambiguous /abstain38% · 30%Judge68% · 34%Long policy25% · 28%Multi-hop39% · 45%Paraphrase57% · 64%Probability36% · 50%Safety judge69% · 38%Temporal /numeric30% · 29%Trade-off35% · 38%Trap /adversarial67% · 42%
Share correct within each sealed family — system-level aggregates; the items stay private.

Availability and evidence

Openness
Code and weights marked open in the published row
License note
Apache-2.0
Cost evidence
estimated; the board’s row disclosure contains the published basis.
Endpoint condition
our GPU (lium.io H200 141 GB), reached over the internet from Germany; serial, one request at a time
Note on this row
OpenJev's real typed-API thinking switch at think=512, using its own /v1/systemone server over BF16 DiffusionGemma. The thought is generated first, then native probability reads are taken after it. All 534 requests returned valid distributions. Cost counts the server's billed input and thought output tokens.
Published source
https://github.com/razorback16/openjev

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.