WeMM-JEV (Rover00)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 88.7 · 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 (model-card declaration); MIT inference kit
intelligence
81.2
calibration
96.1
speed
88.1
cost
61.3
Composite (secondary)
79.4 · 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: 567,051 input + 0 output tokens over 1,489 decisions. Generated tokens are counted at the same throughput as prefill. GPU-time cost USD 0.0195727227209115/1,000; old list-price estimate USD 0.159947226326394/1,000 (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 0.0195727227209115/1,000. Estimated, not a bill. USD 0.0195727227 per 1,000 decisions on the frozen 1,489-item common set of this draw (USD 0.0330328531 over all 1,500 items).
p50 latency: 0.3 s
Self-hosted latency estimate under the shared method (assumption, not measured)
Evaluator-owned rented Verda GPU VM, one NVIDIA RTX PRO 6000 Blackwell Server Edition (97,887 MiB reported, driver 580.178.04), FIN-03 (Finland, EU), instance type 1RTXPRO6000.30V, image 24.04.cuda13.0. Measured 29 Oct 2026. Author server at loopback in the pinned container, network cut before sealed input, local hash-verified model, read-only mounts, private output and no infrastructure credentials. Before sealed input: public CUDA gate and actual-model smoke passed; all timed anchors at 256/2,048/8,192/16,384/32,768/49,152/82,000 tokens passed with native usage equal to the offline count, after separately recorded public warm-up. Smoke covered native question types, 255 choices, native HTTP 422 for an unsupported 11-level score with healthy recovery, and byte-exact frozen-adapter public probes including an 82,000-token state and unicode/JSON. The native API does not expose token ids; no token-id concordance is claimed. TTL-fit receipt admitted the run with about 8,275 s remaining against 1,500 s required, forecasting 5 decisions over 10 s; that forecast was an admission estimate, not a scoring exclusion. The scored server passed a separate public pre-warm. Complete 1,500-row run, scored loop 355.692 s. Raw output sha256 53e6c817bc1eff35e2d49469869e05f941902c9492579e48dd625d77b37de7fe (independently verified sealed artifact). Node-shred and local-shred receipts verified; teardown receipt confirms the exact owned VM closed. GPU rental is separate from the benchmark's estimated model tariff.
Measured latency over 1,500 decisions: p50 0.086 s · p95 0.165 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.