classifier.dev (fast tier)v1.7 · 10 s rule
Data: JevBench v1.8.7, published 2026-11-03 · method v1.7 (10 s rule), re-scored from the recorded per-decision times
Capability 78.2 · API offering, listed and never ranked (wrapper), see the API leaderboard
Measured 2026-10-06 on A4 ∪ P, 600 items, equated (+2.57 I / +3.35 C); round score-a4-2.
MIT (code); hosted service
intelligence
65.2
calibration
91.3
speed
78.2
cost
58.9
Composite (secondary)
71.3 · Not ranked
Cost and measurement conditions
$0.023 per 1,000 decisions (estimated)
v1.5.4 published cost per 1,000 decisions carried (pricing rules unchanged; v1.6 item lengths differ): ESTIMATE (proxy tokens, I-2): operator list price: higher of 19 Sep plan cost and 26 Sep usage tariff USD 0.042/M input (rule 1.2); no exact base-model floor applies
p50 latency: 1.0 s
none (API)
Measured latency over 600 decisions: p50 0.980 s · p95 1.542 s
0 decisions over 10 s.
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.
JevBench v1.6.43 values (method v1.6, before the 10 s rule)
Capability 78.25 · Composite 71.62
intelligence 65.16 · calibration 91.33 · speed 79.54 · cost 58.89
v1.6 speed basis (124 items): p50 0.806 s · p95 1.381 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.
Breakdowns (not part of the score)
Chance-corrected competence (0 = chance, 100 = perfect) on A4 ∪ P (600 items: 300 open + 300 sealed; fewer sealed items than the full set, so wider uncertainty). Raw values, not equated. Category cells use A4+P+L1+L2+L3; per-type/tier cells use the original basis above. Radars against Jev.
| Request type | Open | Sealed |
|---|---|---|
| choice | 79.3 | 74.1 |
| noul | 57.9 | 47.2 |
| score | 66.1 | 50.9 |
Capability by subject topic
| Rules, policy & law | 42.0 | n=1133 |
|---|---|---|
| Coding & software | 62.3 | n=202 |
| Math & numbers | 0.0 | n=269 |
| Finance & commerce | 39.1 | n=178 |
| Support & operations | 57.7 | n=109 |
| Everyday language | 84.2 | n=130 |
| Safety & security | 71.4 | n=84 |
Use cases (TypeSafe categories)
| Model routing | 73.2 | n=267 |
|---|---|---|
| Legal & compliance | 53.9 | n=269 |
| Customer support | 49.8 | n=156 |
| E-commerce | 43.1 | n=88 |
| Insurance claims | 35.1 | n=92 |
| Risk assessment | 29.2 | n=88 |
| Financial crime | 28.9 | n=76 |
| Feature extraction | 23.6 | n=90 |
| Lead generation | 46.3 | n=86 |
| Recruiting | 0.0 | n=88 |
| Knowledge graphs | 33.4 | n=82 |
| LLM guardrails | 79.7 | n=82 |
| Moderation | 62.2 | n=69 |
| Code linting | 32.8 | n=81 |
| Search & retrieval | 84.5 | n=84 |
| Science | 22.9 | n=82 |
| Advertising | 31.1 | n=85 |
| Gaming | 7.5 | n=79 |
| Demand forecasting | 0.0 | n=72 |
Cells need at least 15 answered items; † fewer than 30 (indicative only, listed below the radar in the compare view).
Previous measurement (v1.5.x, not comparable)
Capability 78.2 · composite 71.3 · cost $0.023 per 1,000 decisions (estimated)
Last measured 2026-10-06, measurement release v1.6.1. Older method and scale; the current figures above replace it.