Open source · 28 October 2026
Available nowvLLM Semantic Router launches d3-ultra 27B multimodal decision model
The d3-ultra model reads text, images and video once, then returns probabilities for every answer, scoring 87.2 on the team’s SR‑Text test. It is a 27 B open‑weight multimodal decision model.
Benchmarks are the company's own; not independently verified.
Why it matters
D3-ultra’s 27 B multimodal model returns answer probabilities; consider adding such decision features while noting the internal benchmark.
For product managers
For your product
What does this mean for your product?
Try it on your product More from this issueAll AI news from 28 Oct →
Get this every morning, written for your product.
Deltaday reads the whole AI news cycle and tells you what changed, whether it's real yet, and what to do about it.