Seb-9B-v2: a 9B decision model for comparison (and a router result)
Congratulations on Decision 3.0. Disclosure: I'm the author of Seb-9B-v2 (https://huggingface.co/ironbcc/Seb-9B-v2), a 9B decision model on Qwen3.5-9B, Apache-2.0. It returns a probability for every option in one forward pass, speaks /v1/systemone and a plain chat API, reads images, and takes inputs up to 32k tokens. These are self-run numbers, not independent ones: Decision Index 0.3 public index 59.16 (board numbers: Jev 1.13 57.96, Jev 1.14 60.70), ahead of Jev 1.13 on 22 of 37 benchmarks and clearly behind it on knowledge and reasoning. About 22 ms per decision on one RTX PRO 6000, about 88 ms on an Apple M5 Max. Write-up with the misses included: https://huggingface.co/blog/ironbcc/seb-9b-v2
One result that may interest the semantic-router crowd: using Seb's own confidence as a first tier in front of Claude Opus (420 public decision rows), it answered 51% of requests itself at the same accuracy as Opus alone, at an estimated 51% lower cost. Would you be open to running d3-flash on the same rows? I can share them.