Skip to main content
Create Agent
curl --request POST \
  --url https://api.example.com/labs.chonkie.ai/api/v1/agents
{
  "slug": "<string>",
  "agent": {}
}
Create a new AI agent with customizable LLM settings, system prompts, and knowledge base integration.

​
Examples

import requests

url = "https://labs.chonkie.ai/api/v1/agents"
headers = {
    "Authorization": "Bearer YOUR_API_KEY",
    "Content-Type": "application/json"
}

data = {
    "name": "Documentation Assistant",
    "description": "Helps users navigate product documentation",
    "model": "gpt-4o-mini",
    "systemPrompt": "You are a helpful documentation assistant. Answer questions clearly and concisely.",
    "temperature": 0.3,
    "maxContextChunks": 5,
    "useContext": True,
    "knowledge": "product-documentation"
}

response = requests.post(url, headers=headers, json=data)
result = response.json()

print(f"Created agent: {result['agent']['name']}")
print(f"Slug: {result['slug']}")

​
Request

​
Parameters

​
name
string
required
Agent name (3-100 characters).
​
description
string
Description of the agent’s purpose and capabilities.
​
model
string
default:"gpt-5-mini"
LLM model to use. We support OpenAI models.
​
systemPrompt
string
System prompt that defines the agent’s behavior and personality.
​
maxContextChunks
integer
default:"5"
Maximum number of knowledge base chunks to include in context.
​
useContext
boolean
default:"false"
Whether to inject knowledge base context into conversations.
​
knowledge
string
Knowledge base slug to use for context (required if useContext is true).

​
Response

​
Returns

​
slug
string
Unique slug for the agent (used in API calls).
​
agent
object
Complete agent configuration object.
Each agent object contains the following fields Basic Information
​
id
string
Unique agent identifier.
​
slug
string
URL-friendly unique identifier for the agent.
​
name
string
Display name of the agent.
​
description
string
Description of the agent’s purpose and capabilities.
​
status
string
Agent status (active, inactive, etc.).
Model Configuration
​
model
string
The LLM model used by the agent.
​
systemPrompt
string
System prompt that defines the agent’s behavior and personality.
​
temperature
float
Temperature setting for response randomness (0.0-2.0).
Knowledge Integration
​
useContext
boolean
Whether knowledge base context injection is enabled.
​
knowledge
string
Slug of the connected knowledge base (if any).
​
maxContextChunks
integer
Maximum number of knowledge base chunks to include in context.
Metadata
​
organizationSlug
string
Organization that owns this agent.
​
metadata
object
Additional custom metadata.
​
createdAt
string
Timestamp when the agent was created.
​
updatedAt
string
Timestamp when the agent was last updated.