decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) vs linne-oak 1.0 (31B)
As of Oct 14, 2026, decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) leads linne-oak 1.0 (31B) in 0 of 1 shared categories.
decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted)
Yes. Yes on Decision models (price per 1,000 decisions). Not the overall leader.
linne-oak 1.0 (31B)
Yes. SOTA in co-SOTA in Decision models.
| Category | decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) | linne-oak 1.0 (31B) |
|---|---|---|
| Decision models | #4 · Near-SOTA · Decision models Margin 2.2727272727272663 JevBench · Oct 8, 2026 | #1 · Co-SOTA · Decision models Margin 0 Jev Decision Index · Oct 14, 2026 |
decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted): price per 1M tokens is unset in the catalog.
linne-oak 1.0 (31B): price per 1M tokens is unset in the catalog.
decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted): open weights.
linne-oak 1.0 (31B): open weights.
Decision models: As of Oct 14, 2026, linne-oak 1.0 (31B) leads decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) in Decision models at #1 against #4.
decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) · linne-oak 1.0 (31B)
decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) or linne-oak 1.0 (31B)?
As of Oct 14, 2026, decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) leads linne-oak 1.0 (31B) in 0 of 1 shared categories.
decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) or linne-oak 1.0 (31B) for Decision models?
As of Oct 14, 2026, linne-oak 1.0 (31B) leads decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) in Decision models at #1 against #4.