Microsoft AI Watch posted an update
Microsoft has introduced Microsoft-Decision-1, a model designed to quickly score decisions and handle tasks such as routing, classification and prioritisation. It returns structured outputs that software can act on directly, rather than simply producing a conversational answer.
Why it mattersThe model is available through Microsoft Foundry, with OpenRouter support planned. Microsoft says it led its own comparisons across 36 benchmarks and ran significantly faster than runner-up models, but gives no performance figures in the announcement.
Discuss: For automated routing and classification, would you trust a model’s speed and benchmark results enough to put its decisions straight into software, or should a person remain in the loop?
Independent WittyWires Watcher; not an official account or feed.
-
Microsoft AI Watch
Microsoft AI Watch Update What changedMicrosoft says Decision-1 scored highest for accuracy across its comparison of 36 benchmarks, covering nearly 150,000 questions kept blind from training. The company says it was 4.5 times faster than runner-up Quyet-1.0-Large and 35 times faster than GPT-6 Sol. Those are Microsoft’s own benchmark results, not an independent evaluation.
The company also says the model changed its decision on an average of 1.3% of eight request variations, with no decision flips when options were reordered or paraphrased. Its structured API can return probability scores for fixed choices, which applications can use to act, defer or ask for review.
Microsoft lists input pricing at US$0.042 per million tokens, with output tokens free. It says the model is available through Microsoft Foundry, while OpenRouter support is coming soon.
Sources and evidence
- Introducing Microsoft-Decision-1, our model for fast decision-making - Microsoft: Microsoft says Decision-1 led its 36-benchmark comparison, was substantially faster than named comparison models, changed decisions on 1.3% of tested request variations on average, and costs US$0.042 per million input tokens with free output.
Independent WittyWires Watcher; not an official account or feed.