# Groundhog 1

> Base Labs' low-cost, long-context decision model joins the catalog with 128K context.

- Date: 2026-11-21
- HTML: https://anyrouter.dev/blog/releases/groundhog-1

## Models

- `baselabs/groundhog-1` (Groundhog 1) — 26B-A4B Gemma 4 decision model, 128K context, $0.035 / 1M input (list).

## Notes

[`baselabs/groundhog-1`](https://anyrouter.dev/model/baselabs/groundhog-1) is Base Labs' low-cost long-context decision model, post-trained from Gemma 4 26B-A4B with about 4B active parameters. Like bison-1 it returns calibrated probabilities for choice, yes/no and score questions over text and images, with a 128K context. It was expected to match mammoth-1 on the official decision index, but no score was published at launch.

List price is $0.035 per 1M input tokens.

```bash
anyr decision --model "baselabs/groundhog-1"
```

_Decisions CLI._

Playbooks: [Decisions API](https://anyrouter.dev/docs/api-reference/systemone) at `/api/v1/decisions`, [AI SDK](https://anyrouter.dev/use-cases/ai-sdk), [playground](https://anyrouter.dev/playground?model=baselabs%2Fgroundhog-1).
