Microsoft-Decision-1
Competes with Cloudflare’s Clef decision models, TypeSafe’s Jev and Perplexity’s decider models, a new “decision model” category in autumn 2026.
Model idMicrosoft-Decision-1
Price$0.042 in per 1M tokens; output free
Checked10 October 2026
No published training cutoff
At least2events since its release on 9 October 2026
0 historic, 0 major. 1 day of news.
Microsoft doesn’t publish a training cutoff, so we count from its release date. The real gap is larger.
How to call it
Where to use Microsoft-Decision-1, with its model id.
| Provider | Model id | Endpoint | Docs |
|---|---|---|---|
| Microsoft Foundry | Microsoft-Decision-1 |
commandline.microsoft.com |
Pricing
| Input | $0.042 per 1M tokens |
|---|---|
| Output | $0 per 1M tokens |
As published: “per 1M tokens (USD); output tokens free”.
Source: commandline.microsoft.com, checked 10 October 2026.
Specs
- Input
- Text
- Output
- Text
- Context window
- Not published
- Max output
- Not published
- Open weights
- No
- Licence
- Proprietary
- Training cutoff
- Not published
- Released
- 9 October 2026
- Status
- Current
- Type
- Language model
What stands out
What its maker marketed at launch, and what people found later, each with its source.
Fast decision scoring
A post-trained Qwen3.5-9B that scores choices rather than chatting; Microsoft claims the highest accuracy in a 36-benchmark comparison (~150K questions) and p50 latency 2.5x faster than H2O-Lightning-4B v1.1 and 35x faster than GPT-6 Sol.
Notes
Microsoft says it is also on OpenRouter, but the OpenRouter page returned 404 and the model was absent from OpenRouter’s models API on 2026-10-10. Microsoft plans to rebase it on other models, including Microsoft AI and OpenAI models. No weights release mentioned. Context window not stated.