Jev 1.13: benchmark results
TypeSafe's Jev 1.13 (API model ID jev-1.13.0), the first System One model: it takes natural-language state and returns typed, probabilistic decisions through the System One API instead of generating text. The publisher lists it separately from Jev 1.14. Provider: Other. Released 2026-09-15. Access: API.
Aggregate Rating 1657 ± 24, rank #297 of 1651 rated models, from 38 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LoomBench (agents solo-accept; Typed Decisions) - Attainment | 88.64 | Fraction of 44 instructed agents meeting their objective, in | 100 |
| LoomBench (agents solo-accept; Typed Decisions) - Borrow | 95.5 | Mean source-normalized Borrow objective score (0-100), 11 ag | 100 |
| LoomBench (agents solo-accept; Typed Decisions) - Grow | 98.86 | Mean source-normalized Grow objective score (0-100), 11 agen | 100 |
| LoomBench (agents solo-accept; Typed Decisions) - Hold | 98.18 | Mean source-normalized Hold objective score (0-100), 11 agen | 100 |
| LoomBench (agents solo-refuse; Typed Decisions) - Attainment | 84.09 | Fraction of 44 instructed agents meeting their objective, in | 100 |
| LoomBench (agents solo-refuse; Typed Decisions) - Borrow | 94.09 | Mean source-normalized Borrow objective score (0-100), 11 ag | 100 |
| LoomBench (agents solo-refuse; Typed Decisions) - Give | 100 | Mean source-normalized Give objective score (0-100), 11 agen | 100 |
| LoomBench (agents solo-refuse; Typed Decisions) - Hold | 98.18 | Mean source-normalized Hold objective score (0-100), 11 agen | 100 |
| LoomBench (area fixed; Typed Decisions) - Borrow | 93.05 | Mean source-normalized Borrow objective score (0-100), 11 ag | 100 |
| LoomBench (default; Typed Decisions) - Borrow | 88.5 | Mean source-normalized Borrow objective score (0-100), 11 ag | 100 |
| LoomBench (default; Typed Decisions) - Give | 71.59 | Mean source-normalized Give objective score (0-100), 11 agen | 100 |
| LoomBench (default; Typed Decisions) - Grow | 77.27 | Mean source-normalized Grow objective score (0-100), 11 agen | 100 |
Interactive version: theaggregate.ai/model?slug=jev-1-13 · How It Works · Data refreshed daily, snapshot 2026-10-19.