DevelopingModels
Microsoft releases Decision-1, a Qwen-based decision model priced at $0.042 per million input tokens
Microsoft published Decision-1 on Oct. 9, a model post-trained from Qwen3.5-9B that returns calibrated probabilities for routing, classification and grading. Microsoft says it is the most accurate across a 36-benchmark, nearly 150,000-question comparison and about 2.5 times faster than the runner-up; these are its own figures. Input costs $0.042 per million tokens, output is free, on Foundry and OpenRouter.
- HEAT
- OUTLETS
- 1
- NEW IN 6H
- +1
- FIRST SEEN
- LAST SIGNAL
Primary source
Microsoftcommandline.microsoft.com/microsoft-decision-1-model-foundry/Coverage · 2 articles, oldest first
Saturday, October 10, 2026