# Microsoft-Decision-1.1

> Microsoft-Decision-1.1 is a decision model for routing, classification, prioritization, verification and workflow control. It returns calibrated confidence scores. It uses the System One /decisions API, not chat.

**ID**: `microsoft/microsoft-decision-1.1`  
**Creator**: microsoft  
**Category**: text  
**Context**: 33K tokens  
**Pricing**: $0.04 / $0 per 1M · billed by your provider (BYOK)  
**Released**: 2026-11-17  
**Web page**: https://anyrouter.dev/model/microsoft/microsoft-decision-1.1

**Input modalities**: text  
**Output modalities**: text  

**Capabilities**: systemone

**Supported parameters**: state, model, questions

## Usage

**Endpoint**: `POST https://anyrouter.dev/api/v1/decisions`

```bash
curl https://anyrouter.dev/api/v1/decisions \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "state": "Help! My payouts have been failing for 3 days.",
  "model": "microsoft/microsoft-decision-1.1",
  "questions": {
    "is_urgent": {
      "type": "noul",
      "instructions": "Does this convey urgency?"
    }
  }
}'
```

Decisions is not chat: send `state` and typed `questions` (noul / choice / score). Requires a BYOK key for TypeSafe, LLM Gateway, Vercel AI Gateway, AIHubMix, AutoJev, or Fastino.

## Providers

| Provider | Input | Output |
| --- | --- | --- |
| azure-byok | $0.04/M (list) | $0 (list) |
| vercel-byok | $0.04/M (list) | $0 (list) |

BYOK routes are billed by your own provider key — AnyRouter charges $0. Rates marked (list) are the provider's published list price.

## Model card

- Total parameters: 26B [1]
- Active parameters: 4B [1]

### Strengths [1]

- More than twice the speed of Microsoft-Decision-1 at the same cost
- Highest average accuracy of eight decision models in Microsoft's blind benchmark set
- Returns calibrated confidence so applications can defer when uncertain

### Known limits [2]

- Scores can change with how questions and options are worded or ordered
- Gives no written rationale, and calibration is strongest on familiar tasks
- May reflect base-model bias; not a sole basis for decisions about people

### Use cases [1]

- Routing and classification in software workflows
- Prioritization, verification and workflow control
- Grading model output against a rubric

### API notes [2]

- System One request shape - a state (text or JSON) plus named typed questions
- Answers come back keyed by question name, with token usage

### Links

- announcement: https://commandline.microsoft.com/microsoft-decision-model-ga/
- docs: https://learn.microsoft.com/en-us/azure/foundry/foundry-models/how-to/use-foundry-models-microsoft-decision

### Sources

1. [Microsoft-Decision-1.1 is now generally available](https://commandline.microsoft.com/microsoft-decision-model-ga/) (accessed 2026-11-18)
2. [Deploy and use Microsoft decision models in Microsoft Foundry](https://learn.microsoft.com/en-us/azure/foundry/foundry-models/how-to/use-foundry-models-microsoft-decision) (accessed 2026-11-18)

## Source

- Upstream docs: https://commandline.microsoft.com/microsoft-decision-model-ga/
