shisa-de-1 (shisa-ai, Gemma-4-26B-A4B)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 65.5 · rank #86 of 217 on the open-weights board
Last measured: 2026-10-16 · measurement release: v1.6.26
Apache-2.0
intelligence
49.9
calibration
81.2
speed
92.7
cost
71.1
Composite (secondary)
69.7 · Composite rank #64
Cost and measurement conditions
$0.0092 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: moe gemma, 4B active parameters; throughput 34662.5 tokens/s (architecture-class calibration, 29 Oct 2026, efficient-engine prefill; capped at 100,000 tokens/s). The row's own scored cost basis: 779,620 input + 0 output tokens over 1,485 decisions = 524.996632997 tokens/decision. GPU-time cost USD 0.00915669088398183/1,000; old list-price cost USD 0.047249696969697/1,000; applied min(old list-price cost, GPU-time cost) = USD 0.00915669088398183/1,000. The GPU-time estimate is lower. Estimated, not a bill. Intelligence, Calibration, measured latency and the common cost-item membership are unchanged. Prior prices remain in the archived v1.7 releases.
Measured p50 latency: 0.032 s (on-GPU, no network, loopback; not directly comparable with API latency, which we time end to end from our measuring host, network included)
Ranking and cap use: 0.21 s (est. incl. network (2x + 150 ms))
Self-hosted model: measured p50 0.032 s on-GPU, no network (timed on our GPU over loopback). Not directly comparable with API latency, which we time end to end from our measuring host, network included. The ranking and latency cap use the same estimate for every self-hosted row: 2 × measured + 150 ms = 0.21 s. Model authors can ask us to time a hosted endpoint instead (on request).
x2 + 0.15 s (assumption, not measured)
Evaluator-owned rented Lium GPU pod (measurement receipts on file), 1x NVIDIA H100 80 GB HBM3 (sm_90, driver 580.178.04, United States node, pre-registration Amendments 1-2) - the card most of these systems were first evaluated on, but a different pod, stack and pin set from the earlier runs, so Speed is as measured on this card and probabilities may differ from the earlier runs in the last digits. Python packaging: uv venv from a hashed lock at the old pins, dependencies resolved as of the old build time where the old freeze is lost (declared deviation, UNVERIFIED where noted). Isolation: author server started inside a network namespace with loopback only, the cut proven before and after the sealed input was staged, synthetic smoke gate passed in between; runner unchanged (run_v15.py --resume --resume-keep-failures); one use of input-selfhosted-1500 (under the pre-registration). Published Speed uses the standing x2 + 0.15 s self-hosted adjustment. Stock vLLM server read through the evaluator's shisa_vllm.py adapter.
Measured latency over 1,500 decisions: p50 0.032 s · p95 0.051 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.
JevBench v1.6.43 values (method v1.6, before the 10 s rule)
Capability 65.55 · Composite 62.43 · rank #54 on the all-systems board · #32 open weights
intelligence 49.90 · calibration 81.20 · speed 92.82 · cost 49.77
v1.6 speed basis (124 items): p50 0.031 s · p95 0.049 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.