bison-1 and groundhog-1 are now available through Baseten Model APIs, and the weights for bison-1 and groundhog-1 are open source on Hugging Face. They're decision models that classify inputs, route requests, and score content against a rubric, with typed answers your application can use directly. Each accepts up to 131,072 tokens per question, eight times as many as earlier decision models. bison-1 is post-trained from Gemma 4 31B, and groundhog-1 is post-trained from Gemma 4 26B A4B.
Call decision-models/baseten-bison-latest or decision-models/baseten-groundhog-latest through the System One endpoint with your Baseten API key:
import os
import requests
resp = requests.post(
"https://inference.baseten.co/v1/systemone",
headers={"Authorization": f"Bearer {os.environ['BASETEN_API_KEY']}"},
json={
"model": "decision-models/baseten-groundhog-latest",
"state": "I was charged twice for my subscription. Please refund the duplicate charge.",
"questions": {
"route": {
"type": "choice",
"instructions": "Which team should handle this request?",
"criteria": {
"billing": "Billing, payments, and refunds",
"technical_support": "Technical troubleshooting",
},
},
"refund": {
"type": "noul",
"instructions": "Is the customer asking for a refund?",
},
},
},
)
print(resp.text)Or send the same kind of request to bison-1 with cURL:
curl https://inference.baseten.co/v1/systemone \
-H "Authorization: Bearer $BASETEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "decision-models/baseten-bison-latest",
"state": "The checkout page shows a blank screen after I enter my card details.",
"questions": {
"route": {
"type": "choice",
"instructions": "Which team should handle this request?",
"criteria": {
"billing": "Billing, payments, and refunds",
"technical_support": "Technical troubleshooting"
}
},
"urgency": {
"type": "score",
"instructions": "How urgent is the request?",
"criteria": ["Routine request", "Needs attention today", "Immediate emergency"]
}
}
}'For more information, see our docs.