Unverified50% confidenceFactExact time
Microsoft introduced the Decision-1 model, which does not generate natural language and only outputs probability values: yes/no judgments output noul, multiple-choice judgments output choice, and ordered ratings output score, consuming about 1 output token per call.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/13/2026
First Seen
Valid until: 1/11/2027
Sources
Related Entities
Related Claims
UnverifiedMicrosoft claims that Decision-1 outperforms general-purpose large language models and other specialized decision models on structured decision-making tasks across both the latency and quality dimensions.73% similarUnverifiedAchint Srivastava, Corporate Vice President at Microsoft's Chief Technology Officer office, described the decision model as specifically designed to deliver structured outputs immediately executable by software, distinct from large language models that generate text or reason.68% similarUnverifiedMicrosoft released its first reasoning model built from scratch, MAI-Thinking-1.68% similarUnverifiedMicrosoft CEO Satya Nadella launched the decision model Microsoft-Decision-1, which is based on post-training of Qwen3.5-9B and is specifically designed to handle routing, classification, priority judgment, validation, and workflow control.67% similarUnverifiedDecision models do not generate text; instead, given an input state and a set of named questions, they return typed answers with probabilities in a single forward pass.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1018523API
curl https://kongchang.com/api/v1/knowledge/claims/1018523MCP
get_claim(id=1018523)