# SubQ Snap 1.2

> SubQ Snap 1.2 is Subquadratic's long-context decision model for enterprise decisions. It returns calibrated probabilities for yes/no, choice and score questions over text and images, with a 384K-token context and nearly order-invariant answers. It is BYOK through SubQ's own API.

**ID**: `subquadratic/subq-snap-1.2`  
**Creator**: subquadratic  
**Category**: text  
**Context**: 393K tokens  
**Pricing**: $0.04 / $0 per 1M · billed by your provider (BYOK)  
**Released**: 2026-11-11  
**Web page**: https://anyrouter.dev/model/subquadratic/subq-snap-1.2

**Input modalities**: text, image  
**Output modalities**: text  

**Capabilities**: systemone

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

**Aliases**: `subquadratic/subq-snap-1-2`, `subquadratic/subq-snap-latest`

## 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": "subquadratic/subq-snap-1.2",
  "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 |
| --- | --- | --- |
| subq-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

- Native context: 393216 tokens [3,2]

### Strengths [2,3]

- Tied for first on the independent decision index v2.1 with a score of 73
- Nearly invariant to answer-option ordering, with negligible TTFT overhead
- Scores 68.4 on Decision LCR v1.1 and 69.05 on the open-source component of the Hugging Face Decision Index

### Known limits [1]

- Answers fixed yes/no, choice or score questions; it does not generate chat text
- Image input only, not image output

### Use cases [2]

- Long-document processing and enterprise decisions
- Agentic traces and tool selection
- Policy checks, retrieval and intent classification

### API notes [1,3]

- System One envelope with a state plus typed questions
- Text and image input
- 384K-token context and $0.04 per 1M input tokens, with free output

### Links

- announcement: https://x.com/subquadratic/status/2120641614305517966
- docs: https://docs.subq.ai/using-the-api/models/
- pricing: https://docs.subq.ai/using-the-api/models/
- Previous version: https://anyrouter.dev/model/subquadratic/subq-snap-preview

### Sources

1. [SubQ models](https://docs.subq.ai/using-the-api/models/) (accessed 2026-11-12)
2. [SubQ Snap 1.2: Order-Robust Decisions and Higher Long-Context Accuracy](https://subq.ai/subq-snap-1-2) (accessed 2026-11-12)
3. [SubQ Snap 1.2 announcement](https://x.com/subquadratic/status/2120641614305517966) (accessed 2026-11-12)

## Source

- Upstream docs: https://docs.subq.ai/using-the-api/models/