Microsoft AI Releases Microsoft-Decision-1: A Qwen3.5-9B Decision-Scoring Model
Microsoft posts Qwen3.5-9B as a decision model for agent routing and control—calibrated probability scoring instead of generated text.
Why it matters
Microsoft releases Microsoft-Decision-1, a decision-scoring model purpose-built for agent control, routing, and classification workflows. Available now in Microsoft Foundry and OpenRouter, it represents a shift toward specialized model architectures for deterministic agent decision-making—addressing a real operational need (routing and verification) rather than generic text generation.
The key facts
6 to knowProduct: Microsoft-Decision-1, a decision-scoring model
Base model: Alibaba Qwen3.5-9B (post-trained)
Output: calibrated probability for fixed answer options, not generated text
Use cases: routing, classification, verification, agent control
Availability: Microsoft Foundry and OpenRouter (GA status not explicitly stated but 'available now' implied)
Existing story key match: microsoft-decision-1-ai-gateway-launch
The story so far
Earlier coverage of this storyline
- Microsoft Decision-1 now available on AI GatewayVercel Blog
- This story
Go to the source
MarkTechPostmarktechpost.com
Publisher excerpt: Microsoft has released Microsoft-Decision-1, a decision model for routing, classification, verification and agent control. Microsoft-Decision-1 is a decision-scoring model that returns a calibrated probability for each fixed answer option instead of generated text. It is post-trained from Alibaba’s…