wity-1 (Wity, reasoning always)v1.7 · 10 s rule
Data: JevBench v1.8.18, published 2026-11-20 · method v1.7 (10 s rule), re-scored from the recorded per-decision times
Capability 77.1 · API offering, listed but not ranked on the API leaderboard
registry ranked=false or no composite
Last measured: 2026-10-06 · measurement release: v1.6.1
proprietary hosted API
intelligence
67.1
calibration
87.1
speed
68.5
cost
58.8
Composite (secondary)
69.0 · Not ranked on this board
Cost and measurement conditions
$0.024 per 1,000 decisions (public list price)
Wity's stated API tariff USD 0.042 per 1M input tokens, output free; measured on the v1.6.1 common cost basis (all items except those outside the Jev reference input range) from this run's token usage
p50 latency: 1.9 s
none (API)
operator's hosted production API, measured as is; API default is off, our request sets reasoning=always; labelled configuration variant.
Measured latency over 1,500 decisions: p50 1.907 s · p95 7.387 s
36 decisions over 10 s (36 of them answered without error, counted as wrong).
A decision that takes longer than 10 seconds (measured time) counts as a wrong answer and as zero calibration for that item: each type's calibration is multiplied by the share of its decisions answered within 10 s. The limit uses the measured time (raw). The displayed adjusted time for self-hosted rows (2 × + 0.15 s) is an assumption and is not used for the limit.
Breakdowns (not part of the score)
Chance-corrected competence (0 = chance, 100 = perfect) by category on S+P+L1+L2+L3. Raw values, not equated; the same cells feed the compare view.
Capability by subject topic
| Rules, policy & law | 50.2 | n=1616 |
|---|---|---|
| Coding & software | 67.6 | n=382 |
| Math & numbers | 48.3 | n=370 |
| Finance & commerce | 51.9 | n=223 |
| Support & operations | 41.1 | n=160 |
| Everyday language | 86.1 | n=145 |
| Safety & security | 62.6 | n=109 |
Use cases (TypeSafe categories)
| Model routing | 72.2 | n=524 |
|---|---|---|
| Legal & compliance | 66.9 | n=515 |
| Customer support | 55.4 | n=240 |
| E-commerce | 45.2 | n=119 |
| Insurance claims | 48.8 | n=119 |
| Risk assessment | 70.7 | n=118 |
| Financial crime | 48.5 | n=101 |
| Feature extraction | 53.9 | n=99 |
| Lead generation | 50.6 | n=95 |
| Recruiting | 21.9 | n=94 |
| Knowledge graphs | 10.3 | n=94 |
| LLM guardrails | 34.0 | n=93 |
| Moderation | 58.2 | n=93 |
| Code linting | 39.6 | n=90 |
| Search & retrieval | 72.0 | n=90 |
| Science | 22.2 | n=90 |
| Advertising | 34.2 | n=88 |
| Gaming | 11.6 | n=86 |
| Demand forecasting | 50.4 | n=83 |
Cells need at least 30 answered items; † fewer than 30 (indicative only; drawn as hollow, lighter points on the compare radars).
JevBench v1.6.43 values (method v1.6, before the 10 s rule)
Capability 79.25 · Composite 70.56
intelligence 69.25 · calibration 89.25 · speed 71.20 · cost 58.83
v1.6 speed basis (124 items): p50 1.689 s · p95 4.489 s
These figures are kept for comparison. The current figures above are re-scored under the 10 s rule from the recorded per-decision times and replace them.