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

Qwen3.8 27B (Chutes TEE)

Instruction model, JSON schema · by Qwen / Chutes · API endpoint saw sealed item text

JevBench v1.4.2.2 score

0.0

Partial run, not ranked — the published run is incomplete.

Shown, not ranked. The marked tick is Jev 1.13.0 (63.3).

Published axes

intelligence
40.4
calibration
93.6
speed
61.3
cost
0.0

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: Qwen3.8 27B — Instruction model, JSON schema · Score 0.0 (partial run, not ranked)
  • 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, Qwen3.8 27B vs Jev 1.13.0. Intelligence: 40.4 vs 53.1; Calibration: 93.6 vs 76.3; Speed: 61.3 vs 83.3; Cost: 0.0 vs 52.0.50100Intelligence40.4 · 53.1Calibration93.6 · 76.3Speed61.3 · 83.3Cost0.0 · 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, Qwen3.8 27B vs Jev 1.13.0. Easy: 99% vs 100%; Standard: 99% vs 99%; Judge: 95% vs 95%; Hard: 21% vs 74%; Sealed: 22% vs 37%.50100Easy99% · 100%Standard99% · 99%Judge95% · 95%Hard21% · 74%Sealed22% · 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), Qwen3.8 27B vs Jev 1.13.0. Ambiguous / abstain: 71% vs 43%; Judge: 51% vs 54%; Long policy: 15% vs 44%; Multi-hop: 7% vs 64%; Probability: 21% vs 63%; Temporal / numeric: 8% vs 28%; Trade-off: 16% vs 55%; Routing: 0% vs 100%; Trap / adversarial: 0% vs 83%; Paraphrase: 0% vs 64%; Safety judge: 0% vs 38%.50100Ambiguous /abstain71% · 43%Judge51% · 54%Long policy15% · 44%Multi-hop7% · 64%Probability21% · 63%Temporal /numeric8% · 28%Trade-off16% · 55%Routing0% · 100%Trap /adversarial0% · 83%Paraphrase0% · 64%Safety judge0% · 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, Qwen3.8 27B vs Jev 1.13.0. Ambiguous / abstain: 78% vs 30%; Judge: 93% vs 34%; Long policy: 0% vs 28%; Multi-hop: 0% vs 45%; Paraphrase: 0% vs 64%; Probability: 0% vs 50%; Safety judge: 0% vs 38%; Temporal / numeric: 0% vs 29%; Trade-off: 0% vs 38%; Trap / adversarial: 0% vs 42%.50100Ambiguous /abstain78% · 30%Judge93% · 34%Long policy0% · 28%Multi-hop0% · 45%Paraphrase0% · 64%Probability0% · 50%Safety judge0% · 38%Temporal /numeric0% · 29%Trade-off0% · 38%Trap /adversarial0% · 42%
Share correct within each sealed family — system-level aggregates; the items stay private.

Availability and evidence

Openness
Weights marked open in the published row
License note
open weights
Cost evidence
estimated; the board’s row disclosure contains the published basis.
Endpoint condition
Chutes shared inference (TEE)
Note on this row
69/308 sealed items answered validly (failures count as wrong); partial: Chutes rate limit stopped the run after 81/308 items; unranked as in v1.3

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.