Make.com Workflow Integration

This guide details how to connect KoboiLLM to your Make.com (formerly Integromat) automation scenarios using the standard HTTP - Make a request module. Since KoboiLLM is fully OpenAI-Compatible, you can easily issue HTTP POST requests directly to our endpoints from within Make.com.

---

Step 1: Create an HTTP Module in Make.com

1. Open your Make.com dashboard and create a new scenario.
2. Click the Add a module (+) button, search for HTTP, and select the HTTP app.
3. Select the Make a request action block.

---

Step 2: Configure HTTP Request Parameters

Inside the HTTP - Make a request module configuration dialog, fill in the following parameters:

• URL: https://api.koboillm.com/v1/chat/completions
  (This is the standard OpenAI-Compatible Chat Completions endpoint)
• Method: POST
• Headers: Add these 2 headers to authorize and format your request:
  1. Header 1:
     • Name: Content-Type
     • Value: application/json
  2. Header 2:
     • Name: Authorization
     • Value: Bearer sk-xxxxxx (Replace sk-xxxxxx with your active KoboiLLM Virtual Key)
• Body type: Raw
• Content type: JSON (application/json)
• Request content (JSON Payload): Enter the JSON structure below:

{
  "model": "gemini-3.5-flash",
  "messages": [
    {
      "role": "system",
      "content": "You are a helpful and concise AI assistant."
    },
    {
      "role": "user",
      "content": "Give me a one-sentence tip for morning productivity."
    }
  ],
  "temperature": 0.7
}

• Parse response: Check Yes so Make.com automatically parses the generated JSON response payload.

---

Step 3: Extract the Generated Response Text

Once you run the module (Run once) and receive a successful 200 response:

1. Click the output inspection bubble on top of the HTTP request module.
2. The AI’s generated response text can be accessed at this JSON path:
   data.choices[1].message.content
   (Note: Make.com uses 1-based indexing for arrays, meaning index 0 in standard OpenAI JSON maps to index 1 in Make.com)
3. You can now drag and map the data.choices[1].message.content variable into subsequent Make.com app blocks (such as Google Sheets, Gmail, Slack, Telegram, databases, etc.).

---

Best Practices

• Dynamic Variable Mapping: You can easily replace the static user "content" string with dynamic variables mapped from previous trigger or action modules in your Make.com scenario.
• Model Selection: You can customize the "model" key to target any model available on your KoboiLLM Virtual Key (e.g., gemini-3.5-flash, gpt-4o-mini, etc.).