Decision-King v1.1 (xsdfasfgsa)v1.7 · 10 s rule
Data: JevBench v1.8.1, published 2026-10-30 · method v1.7 (10 s rule), measured under the rule
Capability 82.2 · Not ranked in the current release
Measured on the 27 Oct 2026 draw, which is not yet equated to the reference scale; stronger anchor runs for that draw are being measured. Shown unranked until then.
Last measured: 2026-10-29 · measurement release: v1.8.1
Apache-2.0 (repository LICENSE and model card); base Qwen/Qwen3.8-27B Apache-2.0
intelligence
75.7
calibration
88.8
speed
82.8
cost
61.1
Composite (secondary)
75.6 · Not ranked on this board
Cost and measurement conditions
$0.020 per 1,000 decisions (estimated)
GPU-time estimate from v1.8.0: one monthly-rented RTX PRO 6000 96 GB, EUR 999/month / 730 hours, ECB USD 1.1133/EUR dated 28 Oct 2026, USD 1.5235/hour; average load 70%. Architecture: dense qwen, 27B active parameters; throughput 11763 tokens/s (architecture-class calibration, 29 Oct 2026, efficient-engine prefill; capped at 100,000 tokens/s). The row's own scored cost basis: 574,788 input + 0 output tokens over 1,489 decisions. Generated tokens are counted at the same throughput as prefill. GPU-time cost USD 0.0198397783397036/1,000; old list-price estimate USD 0.16212959032908/1,000 (qwen/qwen3.8-27b at USD 0.42/M input and USD 0/M output). Applied min(old list-price estimate, GPU-time cost) = USD 0.0198397783397036/1,000. Estimated, not a bill. USD 0.0198397783 per 1,000 decisions on the frozen 1,489-item common set of this draw (USD 0.0333017193 over all 1,500 items).
p50 latency: 0.6 s
Self-hosted latency estimate under the shared method (assumption, not measured)
Evaluator-owned rented Verda GPU VM, one NVIDIA RTX PRO 6000 96 GB, FIN-02 (Finland, EU), non-confidential-computing, instance type 1RTXPRO6000.30V, image 24.04.cuda13.0. Measured 29 Oct 2026. Author's native server on loopback, pinned container and local hash-verified model, read-only mounts, private output, no infrastructure credentials, offline flags and network cut before sealed input. Public smoke passed before sealed input: timed probes at 1,000/2,000/4,000/8,000/16,000/32,768/48,000/84,000/100,000 tokens all returned HTTP 200; the 100,000-token probe took 42.046 s and the 84,000-token maximum across two repetitions was 34.014 s. The 131,073-token refusal probe returned native HTTP 400. The serial burst returned 200 of 200 HTTP 200; a separate concurrent busy probe returned one HTTP 200 and one native HTTP 503, recorded only. Native usage offset was 0. Runtime and tokenizer hash receipts were retained before sealed input. TTL-fit admitted with 8,544.55 s remaining against 2,002.57 s required; the five long items were reserved separately and the forecast was not a scoring exclusion. Complete 1,500-row run; scored loop 524.029495 s. Raw output sha256 101f416ebf72693142cf489a83e33ccf0a8db66d75212485459cbc1dd3831fcc (sealed-artifact checksum supplied with the measurement). Row-completion receipt reports complete and cleanup_verified true; node-shred receipt verified. GPU rental is separate from the benchmark's model-cost estimate; rental spend is not supplied as a benchmark tariff.
Measured latency over 1,500 decisions: p50 0.228 s · p95 0.360 s
5 decisions over 10 s (5 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.