d3-ultra offers a single, open-weight model that handles routing, classification, and visual tasks simultaneously, simplifying AI infrastructure by replacing multiple specialized models with one efficient call.
AI-generated summary. Check the original reporting for context and accuracy.
This is the available feed excerpt, which may be shortened or in the publisher’s original language.
The vLLM Semantic Router team has released d3-ultra, a 27B multimodal decision model under Apache-2.0. The d3-ultra decision model does not generate text. It reads an input once and returns a probability for every answer to every question. For developers, that means routing, classification, visual grounding and video checks from 1 open-weight call. TL;DR Size: 26.09B parameters (dense, including a 0.46B vision encoder); 131,072 tokens of context per question. Runs on: tested on 1 AMD Instinct MI325X with BF16 weights; no official quantized variant is listed yet. Performance: highest SR-Text score of the 6 decision models in the team’s own private test. Best: 97.8 on SR-Text tools and automation (chance-corrected accuracy). Worst: 54.6% on SR-Box (9 domains), 5.0 points below its sibling d3-turbo. What is d3-ultra? d3-ultra is a System One decision model from vLLM Semantic Router that answers typed questions with probabilities. It is the newest member of Decision 3.0, a family that spans 0.6B to 27B. You send a state (text or JSON, plus images or videos) and a set of questions. Each question has a type: Choice: pick 1 of several options. Yes / No: return the probability of yes. Score: rate on a scale. Box: locate something in an image or a video moment. Track: follow a target across a whole video. Images can also sit inside questions and options. A quiz can show a photo, then offer 4 etchings as answers. How does d3-ultra work? The core idea is state once. The model reads the input once per request and answers every question from that single read. The config shows a Qwen3.5-architecture backbone with 64 layers. 48 are linear-attention layers and 16 are full-attention layers. The card does not name its base checkpoint. vLLM’s recipe says the sibling d3 was trained from Qwen3.8-27B. Box answers always return exactly 1 box with a confidence. A confidence near 0 means the target is not in the picture. Questions over 131,072 tokens fail with max_length_
An automated sorting aid based on source trust, keywords, and entities, on a scale of 1–12. High: 9–12; medium: 6–8; low: 1–5. It does not measure urgency, verify reporting, or indicate editorial endorsement.
Baidu’s open-source Unlimited OCR 2 enables efficient, single-request parsing of massive documents on a single GPU, significantly lowering the barrier for large-scale text extraction.
OpenAI’s GPT-6.1 Terra halves agent API costs, making high-volume automation significantly cheaper for businesses.
MIT’s significant presence at Climate Week NYC 2026 highlights its active role in connecting global stakeholders to accelerate climate solutions and spur collective action.