HeyGen Decision 1 overview
Decision example
Help! My payments have been failing for 3 days.
| Question | Result |
|---|---|
| Damaged item | P(yes): 0.0009 |
| Team | billing (confidence 0.9979) |
| Urgency | 0.8946 / 2 (confidence 0.8384) |
Actual model response from Nov 20, 2026.
View request and response
Request
{
"model": "heygen/decision-1",
"state": "Help! My payments have been failing for 3 days.",
"questions": {
"damaged_item": {
"type": "noul",
"instructions": "Does the customer report an item that arrived damaged?"
},
"team": {
"type": "choice",
"instructions": "Which team should handle this request?",
"criteria": {
"returns": "Damaged products, replacements and returns",
"billing": "Payments, invoices and charges",
"technical": "App errors and integrations"
}
},
"urgency": {
"type": "score",
"instructions": "Rate how urgently the request needs attention.",
"criteria": [
"Routine follow-up",
"Customer is blocked",
"Immediate safety risk"
]
}
}
}Response
{
"model": "heygen-decision-1",
"answers": {
"damaged_item": {
"noul": 0.0009110512463758376,
"type": "noul"
},
"team": {
"choice": "billing",
"confidence": 0.9978828523950063,
"probabilities": {
"billing": 0.9985885682633375,
"returns": 0.0003795924574247207,
"technical": 0.001031839279237732
},
"type": "choice"
},
"urgency": {
"confidence": 0.8383778292533606,
"legend": {
"0": "Routine follow-up",
"1": "Customer is blocked",
"2": "Immediate safety risk"
},
"probabilities": {
"0": 0.1065642912067318,
"1": 0.8922518861689072,
"2": 0.0011838226243611031
},
"score": 0.8946195314176294,
"type": "score"
}
},
"usage": {
"input_tokens": 790,
"output_tokens": 0
}
}HeyGen Decision 1 by HeyGen
HeyGen built Decision 1 to judge input with no generated text: one request carries 1 to 400 named questions, each returned as a typed result under answers.
HeyGen Decision 1 specs
- Model ID
heygen/decision-1- Developer
- HeyGen
- Context window
- 131,072 tokens
- Input
- Image, Text
- Output
- Text
- Endpoint
POST https://api.aimlapi.com/v1/decisions
Probabilities you can threshold
Every answer carries a calibrated probability between 0 and 1. Your code decides what happens next, for example by sending low-confidence results to a human reviewer.
All questions in a request are asked about the same shared input. Text goes in as a string, a JSON object or a JSON array. A request accepts up to four images.
Pick the question type that fits the decision
A yes/no question returns the probability of yes. You can add descriptions of what true and false mean for that question. Start there for moderation rules, and write the descriptions in the words of your policy.
A choice question takes 2 to 255 options. It returns the selected key, a confidence value and a probability for each option. A score question takes 2 to 10 ordered levels, lowest first. It returns an expected level that can be fractional, such as 1.34, along with a confidence, a probability per level and the legend as you sent it.
Use cases
Flag user-submitted images and text for moderation with one yes/no question per policy rule.
Check an uploaded ID photo and its accompanying text against several identity criteria in one request.
Route support tickets to a queue by asking a choice question over the ticket text and attached screenshots.
Grade content for QA by scoring it on an ordered scale of quality levels.
Capabilities
- Vision
HeyGen Decision 1 API pricing
| Usage | Price | Unit |
|---|---|---|
| Input | $0.14 | 1M tokens |
| Output | $0 | 1M tokens |
Prices from the AI/ML API price list, checked on Nov 20, 2026.
How to use the HeyGen Decision 1 API
View docs (opens in a new tab)Build with your agent
Give your agent the model details, code and documentation.
1Get your API key
Create API keyexport AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"2Make your first request
curl 'https://api.aimlapi.com/v1/decisions' \
-H "Authorization: Bearer $AIMLAPI_KEY" \
-H 'Content-Type: application/json' \
-d '{
"model": "heygen/decision-1",
"state": "Help! My payments have been failing for 3 days.",
"questions": {
"damaged_item": {
"type": "noul",
"instructions": "Does the customer report an item that arrived damaged?"
},
"team": {
"type": "choice",
"instructions": "Which team should handle this request?",
"criteria": {
"returns": "Damaged products, replacements and returns",
"billing": "Payments, invoices and charges",
"technical": "App errors and integrations"
}
},
"urgency": {
"type": "score",
"instructions": "Rate how urgently the request needs attention.",
"criteria": [
"Routine follow-up",
"Customer is blocked",
"Immediate safety risk"
]
}
}
}'import os
import requests
headers = {
"Authorization": f"Bearer {os.environ['AIMLAPI_KEY']}",
"Content-Type": "application/json",
}
payload = {
"model": "heygen/decision-1",
"state": "Help! My payments have been failing for 3 days.",
"questions": {
"damaged_item": {
"type": "noul",
"instructions": "Does the customer report an item that arrived damaged?"
},
"team": {
"type": "choice",
"instructions": "Which team should handle this request?",
"criteria": {
"returns": "Damaged products, replacements and returns",
"billing": "Payments, invoices and charges",
"technical": "App errors and integrations"
}
},
"urgency": {
"type": "score",
"instructions": "Rate how urgently the request needs attention.",
"criteria": [
"Routine follow-up",
"Customer is blocked",
"Immediate safety risk"
]
}
}
}
response = requests.post("https://api.aimlapi.com/v1/decisions", headers=headers, json=payload, timeout=120)
response.raise_for_status()
print(response.json())const response = await fetch("https://api.aimlapi.com/v1/decisions", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.AIMLAPI_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"model": "heygen/decision-1",
"state": "Help! My payments have been failing for 3 days.",
"questions": {
"damaged_item": {
"type": "noul",
"instructions": "Does the customer report an item that arrived damaged?"
},
"team": {
"type": "choice",
"instructions": "Which team should handle this request?",
"criteria": {
"returns": "Damaged products, replacements and returns",
"billing": "Payments, invoices and charges",
"technical": "App errors and integrations"
}
},
"urgency": {
"type": "score",
"instructions": "Rate how urgently the request needs attention.",
"criteria": [
"Routine follow-up",
"Customer is blocked",
"Immediate safety risk"
]
}
}
}),
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
console.log(await response.json());HeyGen Decision 1 FAQ
How much does HeyGen Decision 1 cost?
HeyGen Decision 1 is billed at $0.14 / 1M tokens for input. There are no output-token charges.
How many images can one request include?
HeyGen Decision 1 accepts up to four images per request. Text can be sent as a string, a JSON object or a JSON array.
What does a yes/no question return?
It returns the probability that the answer is yes, as a value from 0 to 1. You can add descriptions of what true and false mean for that question.
What does a choice question need?
Provide 2 to 255 options as criteria, each with a description or null. The answer is one option key, a confidence value and a probability per option.
How do I define the scale for a score question?
Send 2 to 10 ordered level descriptions, lowest first, as text or a JSON object or array. The response returns the expected level, a confidence, a probability for each level and a legend that maps each level index to the description you sent.
What is the HeyGen Decision 1 model ID?
Use heygen/decision-1 as the model parameter. Requests go to https://api.aimlapi.com/v1/decisions with your AI/ML API key.