Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text) vs decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted)
As of Oct 14, 2026, Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text) leads decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) in 0 of 1 shared categories.
Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text)
#37 Best current rank: #37 in Decision models, as of Oct 14, 2026.
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.
| Category | Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text) | decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) |
|---|---|---|
| Decision models | #37 · #37 in Decision models Margin 16.13636363636364 Jev Decision Index · Oct 14, 2026 | #4 · Near-SOTA · Decision models Margin 2.2727272727272663 JevBench · Oct 8, 2026 |
Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text): 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): price per 1M tokens is unset in the catalog.
Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text): open weights.
decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted): open weights.
Decision models: 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 Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text) in Decision models at #4 against #37.
Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text) · decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted)
Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text) or decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted)?
As of Oct 14, 2026, Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text) leads decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) in 0 of 1 shared categories.
Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text) or decisio v0.8.0 on gemma-4-31B-it (frozen, FP8 on load, one prefill per state, self-hosted) for Decision models?
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 Clef (Cloudflare, Qwen3.8-27B post-train with a joint schema head, multimodal, measured on text) in Decision models at #4 against #37.