AI · Oct 15, 2026
NVIDIA Releases Nemotron 3.5 Super VL, a 120B Open Multimodal Agent ModelLiquid AI Releases d1-ultra: A Multimodal Decision Model That Scores 65 on the AA Decision Index
Oct 15, 2026, 11:26 PM · MarkTechPost
Liquid AI’s d1-ultra returns calibrated probabilities in one forward pass — 65 on Artificial Analysis’s Decision Index at $0.042 per million input tokens.
Why it matters
MarkTechPost reports Liquid AI released d1-ultra, a text-and-image decision model that returns calibrated probabilities in a single forward pass. It scores 65 on the Artificial Analysis Decision Index and costs $0.042 per million input tokens.
From the desk
Decision models are the week’s theme — Jev’s adoption spike, twenty alternatives, now Liquid’s multimodal entry with a published index score and aggressive input pricing.
Calibrated probabilities in one pass are exactly what production routers want. We’ll trust the 65 more once Artificial Analysis’s methodology is cross-checked against other decision models in the same harness. Cheap multimodal decisions could make agent graphs feel snappy instead of essay-slow.
I’m watching head-to-heads with Jev and whether image inputs become table stakes for the category.
Context
The broader market saw TypeSafe’s Jev launch September 15 and a rapid wave of competing decision models within about four weeks, per MarkTechPost.
Who feels it
- Agent developers
- A multimodal, calibrated decision option at low input token cost.
- Decision-model startups
- Another scored competitor on Artificial Analysis’s index.
- Enterprise architecture teams
- More choice for classification hops that shouldn’t call a frontier chat model.
What to watch
- Artificial Analysis comparisons against Jev and peers
- Real-world calibration quality outside the index
- Whether $0.042/M input pricing holds at scale