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

system-one (Qwen3-8B, Sean Goedecke)

Jev rebuild · by Sean Goedecke

JevBench v1.4.2.2 score

23.4

Rank #57 of 91 ranked systems.

39.9 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
43.6
calibration
32.8
speed
84.4
cost
41.5

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: system-one — Jev rebuild · Score 23.4 (#57)
  • 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, system-one vs Jev 1.13.0. Intelligence: 43.6 vs 53.1; Calibration: 32.8 vs 76.3; Speed: 84.4 vs 83.3; Cost: 41.5 vs 52.0.50100Intelligence43.6 · 53.1Calibration32.8 · 76.3Speed84.4 · 83.3Cost41.5 · 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, system-one vs Jev 1.13.0. Easy: 100% vs 100%; Standard: 91% vs 99%; Judge: 92% vs 95%; Hard: 50% vs 74%; Sealed: 24% vs 37%.50100Easy100% · 100%Standard91% · 99%Judge92% · 95%Hard50% · 74%Sealed24% · 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), system-one vs Jev 1.13.0. Ambiguous / abstain: 24% vs 43%; Judge: 43% vs 54%; Long policy: 22% vs 44%; Multi-hop: 34% vs 64%; Probability: 40% vs 63%; Temporal / numeric: 22% vs 28%; Trade-off: 37% vs 55%; Routing: 100% vs 100%; Trap / adversarial: 70% vs 83%; Paraphrase: 29% vs 64%; Safety judge: 31% vs 38%.50100Ambiguous /abstain24% · 43%Judge43% · 54%Long policy22% · 44%Multi-hop34% · 64%Probability40% · 63%Temporal /numeric22% · 28%Trade-off37% · 55%Routing100% · 100%Trap /adversarial70% · 83%Paraphrase29% · 64%Safety judge31% · 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, system-one vs Jev 1.13.0. Ambiguous / abstain: 19% vs 30%; Judge: 29% vs 34%; Long policy: 13% vs 28%; Multi-hop: 21% vs 45%; Paraphrase: 29% vs 64%; Probability: 32% vs 50%; Safety judge: 31% vs 38%; Temporal / numeric: 20% vs 29%; Trade-off: 35% vs 38%; Trap / adversarial: 42% vs 42%.50100Ambiguous /abstain19% · 30%Judge29% · 34%Long policy13% · 28%Multi-hop21% · 45%Paraphrase29% · 64%Probability32% · 50%Safety judge31% · 38%Temporal /numeric20% · 29%Trade-off35% · 38%Trap /adversarial42% · 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
no licence file in the repository as of 19 Sep; Qwen3 weights Apache-2.0
Cost evidence
estimated; the board’s row disclosure contains the published basis.
Endpoint condition
our RunPod GPU (RTX PRO 4500 Blackwell 32 GB (EU-RO-1)), reached over the internet
Published source
https://github.com/sgoedecke/system-one

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.