SPX-CD Pro (SurdAI, hosted /v1/systemone, Oct-4 checkpoint)v1.7 · 10 s rule

Data: JevBench v1.8.19, published 2026-11-21 · method v1.7 (10 s rule), re-scored from the recorded per-decision times

Capability 78.5 · API offering, ranked on the API leaderboard

cost 2.4× Jev (v1.5 reference)

Last measured: 2026-10-06 · measurement release: v1.6.1

Licence not stated

Base model: Qwen/Qwen3.8-27Bmodel card · Vendor model card for the published adapter (9 Oct); the hosted Oct-4 checkpoint measured here is not stated to be identical.

intelligence

64.6

calibration

92.3

speed

77.4

cost

43.1

Composite (secondary)

47.6 · Not ranked on this board

Cost and measurement conditions

$0.079 per 1,000 decisions (estimated)

Operator's own public-beta overage tariff, conservative reading USD 0.08 per 1M input tokens, output USD 0: three dated primary readings of the operator's own published overage tariff: 24 Sep 2026 provider console (flash 0.025, pro 0.08 per 1M input), 1 Oct 2026 docs page (flash 0.04, pro 0.02), 6 Oct 2026 docs page AND /v1/models payload agreeing (flash 0.02, pro 0.04). Sources conflict, so the per-model MAXIMUM is used, which is the published fastino-gliner-2-5-decide precedent ("published sources conflict ... conservative higher"). Today's agreeing pair would be USD 0.04/M, i.e. half this row's cost. No base-model floor applies: the submitter states the base model stays sealed until a later open release. Priced on usage.input_tokens as the provider reports them. The provider also reports billable_input_tokens = input_tokens / 2 on every single item (297/297 of the public set, exactly 2.0): its default effort 2 renders the prompt twice and its docs bundle the second pass free during public beta ("default effort 2 is included at the single-pass input price"). That bundle is a public-beta promotion, and TASK.md forbids a free or promotional tier setting the price, so the two passes actually consumed are charged. The operator's own billed figure is exactly half of this row's cost

p50 latency: 1.1 s

none (API)

measured 2026-10-06 on the live v1.6.0 S u P (1,500 rows) under the published 5 Oct 2026 full-set amendment, same lane, same adapter and same provider default as spx-cd-flash; add-requests run 53, benchmarkheaven.com/submit ref 3FF00545

Measured latency over 1,500 decisions: p50 1.145 s · p95 1.576 s

1 decision over 10 s (1 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 & law57.1n=1616
Coding & software77.9n=382
Math & numbers15.3n=370
Finance & commerce48.0n=223
Support & operations54.9n=160
Everyday language72.9n=145
Safety & security51.7n=109

Use cases (TypeSafe categories)

Model routing77.2n=524
Legal & compliance72.8n=515
Customer support52.2n=240
E-commerce55.6n=119
Insurance claims47.7n=119
Risk assessment52.2n=118
Financial crime30.4n=101
Feature extraction34.0n=99
Lead generation58.8n=95
Recruiting0.2n=94
Knowledge graphs56.1n=94
LLM guardrails78.7n=93
Moderation52.4n=93
Code linting73.5n=90
Search & retrieval88.5n=90
Science24.6n=90
Advertising28.6n=88
Gaming7.0n=86
Demand forecasting0.0n=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 78.64 · Composite 47.67 · rank #68 on the all-systems board · #25 API

intelligence 64.86 · calibration 92.41 · speed 77.96 · cost 43.08

v1.6 speed basis (124 items): p50 1.128 s · p95 1.417 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.

Published source