Integrasi API Preview

API Integration Preview is a feature provided by the Service Portal AI that displays an auto-generated code snippet for each prompt you submit to the model via the Interactive Chat panel. So, every time you test a prompt in the Playground Tab, Deka LLM automatically generates a ready-to-use API call example.

cURL

cURL (Client URL) is a command-line tool used to send requests to servers using various protocols such as HTTP, HTTPS, and FTP. Below is an example of a cURL command generated when you input a prompt.

curl https://dekallm.cloudeka.ai/v1/chat/completions \
  -H "Authorization": "Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "meta/llama-4-maverick-instruct",
    "messages": [{"role": "user", "content": ""}],
    "temperature": 0.6,
    "top_p": 0.7
  }'

Explanation of the cURL Command.

curl https://dekallm.cloudeka.ai/v1/chat/completions \

This line curl https://dekallm.cloudeka.ai/v1/chat/completions \defines the API endpoint URL used by Deka LLM to request a chat completion.

  -H "Authorization": "Bearer YOUR_API_KEY" \

This line -H "Authorization": "Bearer YOUR_API_KEY" \ is used to authenticate the API request by providing the API key in a Bearer token format.

  -H "Content-Type: application/json" \

This line -H "Content-Type: application/json" \ sets the request payload format to JSON.

  -d '{
    "model": "meta/llama-4-maverick-instruct",
    "messages": [{"role": "user", "content": ""}],
    "temperature": 0.6,
    "top_p": 0.7
  }'

This JSON object is sent via an HTTP POST request and contains:

• "model": "meta/llama-4-maverick-instruct",

  Specifies the LLM model to use in Deka LLM.

• "messages": [{"role": "user", "content": ""}],

  An array of messages, where "role": "user" represents the you, and "content": "" is the prompt.

• temperature": 0.6

  Controls the creativity or randomness of the response (higher value = more creative).

• "top_p": 0.7

  Configures nucleus sampling to control the cumulative probability for token selection, affecting the randomness.

Python

Python is a programming language known for its simple syntax, readability, and support for imperative, functional, and object-oriented paradigms. Below is the example code auto-generated when you input a prompt.

from openai import OpenAI

client = OpenAI(
    base_url="https://dekallm.cloudeka.ai/v1",
    api_key="YOUR_API_KEY",
)

completion = client.chat.completions.create(
    model="baai/bge-multilingual-gemma2",
    messages=[{"role": "user", "content": ""}],
    temperature=0.6,
    top_p=0.7,
)
print(completion.choices[0].message.content)

Explanation of the Python Code:

from openai import OpenAI

Imports the OpenAI class from the official OpenAI Python SDK, which provides an interface to interact with LLM API.

client = OpenAI(
    base_url="https://dekallm.cloudeka.ai/v1",
    api_key="YOUR_API_KEY",
)

This line is used to create a client instance to communicate with the Deka LLM API endpoint. There are two important parameters used

• base_url displays the URL of the Deka LLM endpoint,

• api_key is sed to authenticate requests and is taken from the API Key column.

completion = client.chat.completions.create(
    model="meta/llama-4-maverick-instruct",
    messages=[{"role": "user", "content": ""}],
    temperature=0.6,
    top_p=0.7,
)

This line client.chat.completions.create()is used to send a request for chat completion. There are four important parameters used, namely:

• model="meta/llama-4-maverick-instruct",

  Specifies the LLM model to use in Deka LLM.

• messages=[{"role": "user", "content": ""}],

  An array of messages, where "role": "user" represents the you, and "content": "" is the prompt.

• temperatur=0.6

  Controls the creativity or randomness of the response (higher value = more creative).

• and top_p=0.7

  Configures nucleus sampling to control the cumulative probability for token selection, affecting the randomness.

print(completion.choices[0].message.content)

This line print(completion.choices[0].message.content) is used to represent the model's response to the message you send.

Node.js

Node.js is a runtime environment for executing JavaScript code outside the browser. Below is an example Node.js code generated when you input a prompt.

import OpenAI from "openai";

const openai = new OpenAI({
  apiKey: "YOUR_API_KEY",
  baseURL: "https://dekallm.cloudeka.ai/v1",
});

const chatCompletion = await openai.chat.completions.create({
  model: "baai/bge-multilingual-gemma2",
  messages: [{ role: "user", content: "" }],
  temperature: 0.6,
  top_p: 0.7,
});

Explanation of the Node.js Code:

import OpenAI from "openai";

This line import OpenAI from "openai"; imports the OpenAI class from the official Node.js SDK, which allows you to interact with the Deka LLM API.

const openai = new OpenAI({
  apiKey: "YOUR_API_KEY",
  baseURL: "https://dekallm.cloudeka.ai/v1",
});

This line const openai = new OpenAI({...}); is used to create your instance withapiKey andbaseURL configurations.

const chatCompletion = await openai.chat.completions.create({
  model: "baai/bge-multilingual-gemma2",
  messages: [{ role: "user", content: "" }],
  temperature: 0.6,
  top_p: 0.7,
});

This line const chatCompletion = await openai.chat.completions.create({...}); is used to send a chat completion request to the model you are using in Deka LLM with parameters. There are four important parameters used, namely:

• model="meta/llama-4-maverick-instruct",

  Specifies the LLM model to use in Deka LLM.

• messages=[{"role": "user", "content": ""}],

  An array of messages, where "role": "user" represents the you, and "content": "" is the prompt.

• temperatur=0.6

  Controls the creativity or randomness of the response (higher value = more creative).

• and top_p=0.7

  Configures nucleus sampling to control the cumulative probability for token selection, affecting the randomness.