OpenJev (DiffusionGemma 26B-A4B NVFP4, razorback16)
Data: JevBench v1.6.1, published 2026-10-06
Capability 50.2 · Not ranked in the current release
Last measured: 2026-09-27T18:49:51+00:00 · measurement release: v1.4.2.2
Apache-2.0 (repo and weights)
intelligence
45.4
calibration
55.0
speed
83.2
cost
45.5
Composite (secondary)
36.9 · Not ranked on this board
Cost and measurement conditions
$0.066 per 1,000 decisions (estimated)
ESTIMATE: hosted-provider price, openrouter google/gemma-4-26b-a4b-it list price $0.09/M in, $0.3/M out (DiffusionGemma 26B-A4B is not listed; the same-size Gemma 4 26B-A4B MoE sibling is (size class moe_26B-A4B)) x 380 input and 1 output tokens per decision (input tokens measured) [corrected in v1.2.3: the price now averages each of the 314 v1.1 decisions once; see results/v1.2/cost-correction-v1.2.3.json] | ESTIMATE: openrouter google/gemma-4-26b-a4b-it $0.09/M in, $0.3/M out x 1222 in / 0 out tokens per hard decision
Measured p50 latency: 0.241 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: 0.63 s (est. incl. network (2x + 150 ms))
Self-hosted model: measured p50 0.241 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 = 0.63 s. Model authors can ask us to time a hosted endpoint instead (on request).
x2 + 0.15 s (assumption, not measured)
RunPod RTX PRO 4500 Blackwell 32 GB (EU-RO-1)