Private Endpoint

Overview

Private Endpoint offers secure, FADP-compliant access to powerful Large Language Models (LLMs) hosted exclusively in Switzerland — ensuring data privacy and regulatory peace of mind.

Security and Data Privacy

GuardOS is committed to ensuring the security and privacy of your data. See Security and Data Privacy for more information.

Pricing

Pricing is based on expected volume. Please contact us for a quote.

Endpoints

Method	Endpoint
POST	`https://api.guardos.ai/api/v1/private-endpoint`

POST: /api/v1/private-endpoint

Request Headers

interface PrivateEndpointRequest {
	'Content-Type': 'application/json'
	'x-api-key': string
}

Header Name	Type	Description
Content-Type	string	`application/json`
x-api-key	string	Your API key

Request Body

interface PrivateEndpointRequest {
	messages: {
		role: 'user' | 'assistant'
		content: string
	}[]
	settings?: {
		model?: string // Default: deepseek-v3-0324
		stream?: boolean // Default: false
		temperature?: number // Default: 0
		top_p?: number // Default: 1
		top_k?: number // Default: 0
		frequency_penalty?: number // Default: 0
		presence_penalty?: number // Default: 0
		system?: string // Default: You are a helpful assistant.
	}
}

Field Name	Type	Description	Default Value
messages	Message[]	Array of messages
settings	Settings	Optional settings
model	string	Model identifier to use	deepseek-v3-0324
stream	boolean	Stream response	false
temperature	number	Temperature	0
top_p	number	Top P	1
top_k	number	Top K	0
frequency_penalty	number	Frequency penalty	0
presence_penalty	number	Presence penalty	0
system	string	System prompt	You are a helpful assistant.

Success Response

interface PrivateEndpointResponse {
	text: string
	files: []
	reasoningDetails: []
	toolCalls: []
	toolResults: []
	finishReason: 'stop'
	warnings: []
	sources: []
}

Field Name	Type	Description
text	string	Response text of the model
files	[]	Files generated by the model
reasoningDetails	[]	Reasoning details
toolCalls	[]	Tool calls
toolResults	[]	Tool results
finishReason	string	Finish reason
warnings	[]	Warnings
sources	[]	Sources

Error Response

interface PrivateEndpointErrorResponse {
	statusCode: number
	error: string
	message: string
}

Field Name	Type	Description
statusCode	number	Error code
error	string	Error message
message	string	Error message

Examples

Basic Request

For a basic request, you only need to provide the API key and the messages object.

curl

curl -X POST 'https://api.guardos.ai/api/v1/private-endpoint' \
-H 'Content-Type: application/json' \
-H 'x-api-key: YOUR_API_KEY' \
-d '{
  "messages": [
    { "role": "user", "content": "Hello, how are you?" }
  ]
}'

TypeScript

const response = await fetch('https://api.guardos.ai/api/v1/private-endpoint', 
{
    method: 'POST',
    headers: { 
        'Content-Type': 'application/json', 
        'x-api-key': 'YOUR_API_KEY' 
    },
    body: JSON.stringify({ 
        messages: [{ role: "user", content: "Hello, how are you?" }],
    }),
});
const data = await response.json();

Python

import requests
import json

response = requests.post(
    'https://api.guardos.ai/api/v1/private-endpoint',
    headers={'Content-Type': 'application/json', 'x-api-key': 'YOUR_API_KEY'},
    data=json.dumps({'messages': [{'role': 'user', 'content': 'Hello, how are you?'}]})
)
data = response.json()

Basic Streaming Request

For a basic streaming request set settings.stream = true.

curl

curl --no-buffer -X POST https://api.guardos.ai/api/v1/private-endpoint \
  -H "Content-Type: application/json" \
  -H "x-api-key: YOUR_API_KEY" \
  -d '{
    "messages": [{"role": "user", "content": "Hello"}],
    "settings": {
        "stream": true
    }
  }'

TypeScript

const response = await fetch('https://api.guardos.ai/api/v1/private-endpoint', 
{
    method: 'POST',
    headers: { 
        'Content-Type': 'application/json', 
        'x-api-key': 'YOUR_API_KEY' 
    },
    body: JSON.stringify({ 
        messages: [{ role: "user", content: "Hello, how are you?" }],
        settings: {
            stream: true // <-- Enable streaming
        }
    }),
});

const reader = response.body.getReader();
const decoder = new TextDecoder('utf-8');

while (true) {
    const { done, value } = await reader.read();
    if (done) break;
    const chunk = decoder.decode(value, { stream: true });
    console.log(chunk); // Or parse/accumulate it as needed
}

Python

import httpx
import asyncio
import json

async def main():
    url = 'https://api.guardos.ai/api/v1/private-endpoint'
    headers = {
        'Content-Type': 'application/json',
        'x-api-key': 'YOUR_API_KEY',
    }
    data = {
        "messages": [{"role": "user", "content": "Hello"}],
        "settings": {
            "stream": True, # <-- Enable streaming
        }
    }

    async with httpx.AsyncClient() as client:
        async with client.stream("POST", url, headers=headers, json=data) as response:
            async for chunk in response.aiter_text():
                print(chunk, end='')

asyncio.run(main())

Custom system prompt

You can set a custom system prompt by setting settings.system to your desired system prompt.

curl

curl -X POST 'https://api.guardos.ai/api/v1/private-endpoint' \
-H 'Content-Type: application/json' \
-H 'x-api-key: YOUR_API_KEY' \
-d '{
  "messages": [
    { "role": "user", "content": "Hello, how are you?" }
  ],
  "settings": {
    "system": "You are to only respond in German."
  }
}'

TypeScript

const response = await fetch('https://api.guardos.ai/api/v1/private-endpoint', 
{
    method: 'POST',
    headers: { 
        'Content-Type': 'application/json', 
        'x-api-key': 'YOUR_API_KEY' 
    },
    body: JSON.stringify({ 
        messages: [{ role: "user", content: "Hello, how are you?" }],
        settings: {
            system: "You are to only respond in German." // <-- Set custom system prompt
        }
    }),
});
const data = await response.json();

Python

import requests
import json

response = requests.post(
    'https://api.guardos.ai/api/v1/private-endpoint',
    headers={'Content-Type': 'application/json', 'x-api-key': 'YOUR_API_KEY'},
    data=json.dumps({
        'messages': [{'role': 'user', 'content': 'Hello, how are you?'}],
        'settings': {
            'system': "You are to only respond in German." # <-- Set custom system prompt   
        }
    })
)
data = response.json()

With each request you can set custom model settings. For example you can set the temperature to control the randomness of the model's responses. A low value will make the model more deterministic, while a high value will make the model more creative.

Range: 0 - 1.0
Default: 0

curl

curl -X POST 'https://api.guardos.ai/api/v1/private-endpoint' \
-H 'Content-Type: application/json' \
-H 'x-api-key: YOUR_API_KEY' \
-d '{
  "messages": [
    { "role": "user", "content": "Hello, how are you?" }
  ],
  "settings": {
    "temperature": 0.5
  }
}'

TypeScript

const response = await fetch('https://api.guardos.ai/api/v1/private-endpoint', 
{
    method: 'POST',
    headers: { 
        'Content-Type': 'application/json', 
        'x-api-key': 'YOUR_API_KEY' 
    },
    body: JSON.stringify({ 
        messages: [{ role: "user", content: "Hello, how are you?" }],
        settings: {
            temperature: 0.5 // <-- Set custom temperature
        }
    }),
});
const data = await response.json();

Python

import requests
import json

response = requests.post(
    'https://api.guardos.ai/api/v1/private-endpoint',
    headers={'Content-Type': 'application/json', 'x-api-key': 'YOUR_API_KEY'},
    data=json.dumps({
        'messages': [{'role': 'user', 'content': 'Hello, how are you?'}],
        'settings': {
            'temperature': 0.5 # <-- Set custom temperature
        }
    })
)
data = response.json()