- Focused on structured decisions like routing, classification, prioritization and workflow control rather than open-ended generation
- Emits calibrated confidence and can abstain, reducing missteps in critical business flows
- Input priced at $0.042 per million tokens with free output; 4.4x cheaper in internal routing use
AI CURATED HOT STORY · 11/18/2026
Microsoft Launches Decision-1.1 Model with Over 2x Speedup
Microsoft has officially launched Microsoft-Decision-1.1, a specialized model for high-frequency enterprise decision-making scenarios. Instead of generating general text, the model makes structured choices from allowed actions within developer-defined rule constraints, outputs calibrated confidence scores, and can abstain or escalate to humans when information is insufficient. It is now available on Foundry, OpenRouter, and Vercel, priced at $0.042 per million input tokens with free output.
On the Artificial Analysis Decision Index v2.1, Decision-1.1 scores 68, a 6-point improvement over its predecessor, with all six sub-evaluations showing gains, ranking 8th overall. Per-task latency dropped from 274 ms to 123 ms, more than 2x faster, ranking 5th in speed. Internally, for Copilot request routing, it is 2.2x faster and 4.4x cheaper. The model focuses on structured decisions such as routing, classification, prioritization, and workflow control, with calibrated confidence and abstention to reduce misoperation risks in critical business processes.
The release marks an evolution from general-purpose generation to specialized decision-making, offering efficient and reliable support for enterprise automation. Its low pricing and high speed are expected to drive large-scale adoption in customer service, operations, and risk control, intensifying competition in the specialized model segment.
Hourly heat curve
6 fully observed hoursThe line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.
Coverage timeline
Total 2 reports · Latest first- IT之家 · 智能时代T2·78 pts
- Artificial Analysis (@ArtificialAnlys)T1·38 pts
- Scores 68 on Decision Index v2.1, up 6 points from 62 with gains across all six sub-evaluations
- Time per task falls to 123 ms from 274 ms, ranking #5 on speed
- Pricing unchanged at $0.042 per million input tokens; overall rank #8, one point behind d1.2