Images & Vision

ChatboxAI supports both Image Generation (creating images from text) and Vision (understanding images provided to the chat model).

1. Image Generation

Use the images/generations endpoint to generate high-quality images from text prompts.

POST https://chatboxai.co.in/api/v1/images/generations

Supported Models

Test via Postman / Thunder Client (Quick Test)

If you are testing via an API client, make sure to use the prompt field (NOT messages). Image models do not accept chat formats.

• Method: POST
• URL: https://chatboxai.co.in/api/v1/images/generations
• Headers:
  • Authorization: Bearer cbx_live_YOUR_API_KEY
  • Content-Type: application/json
• Body (JSON):
  JSON

  Copy

  {
    "model": "chatboxai/image-flux-schnell",
    "prompt": "A futuristic city skyline at sunset, cyberpunk style, highly detailed 8k.",
    "n": 1,
    "size": "1024x1024"
  }

Example Request (Node.js)

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.CHATBOXAI_API_KEY,
  baseURL: "https://chatboxai.co.in/api/v1",
});

async function generateImage() {
  const response = await client.images.generate({
    model: "chatboxai/image-flux-dev", // Paid, high-quality model
    prompt: "A futuristic city skyline at sunset, cyberpunk style, highly detailed 8k.",
    n: 1,
    size: "1024x1024",
  });

  // The API returns a URL to the generated image
  console.log("Image URL:", response.data[0].url);
}

generateImage();

2. Vision (Image Analysis)

You can pass images into compatible chat models to analyze visual content, read text from images (OCR), or understand charts.

Supported Vision Models

• chatboxai/gemini-2.5-flash (Free)
• chatboxai/gemini-2.5-pro (Paid)
• chatboxai/claude-4.5-sonnet (Paid)
• chatboxai/llama-3.2-90b-vision (Paid)

Passing Images as URLs

You can provide an array of content inside the user message, passing both text and a public image URL.

const response = await client.chat.completions.create({
  model: "chatboxai/gemini-2.5-flash", // Free vision model
  messages: [
    {
      role: "user",
      content: [
        { type: "text", text: "What's in this image?" },
        {
          type: "image_url",
          image_url: {
            url: "https://upload.wikimedia.org/wikipedia/commons/a/a7/React-icon.svg",
          },
        },
      ],
    },
  ],
});

console.log(response.choices[0].message.content); 
// Output: "This is the official logo for React, a JavaScript library..."

Passing Images as Base64

If the image is local (on the user's device or server), you must convert it to a Base64 string.

import fs from "fs";

// 1. Read the image and convert to Base64
const imageBuffer = fs.readFileSync("./chart.png");
const base64Image = imageBuffer.toString("base64");
const mimeType = "image/png";

// 2. Pass to the API
const response = await client.chat.completions.create({
  model: "chatboxai/claude-4.5-sonnet", // High-end paid reasoning
  messages: [
    {
      role: "user",
      content: [
        { type: "text", text: "Explain the trend in this chart." },
        {
          type: "image_url",
          image_url: {
            url: `data:${mimeType};base64,${base64Image}`,
          },
        },
      ],
    },
  ],
});

3. Best Practices for Vision

• Downscale before sending: Very large images consume massive amounts of tokens (and therefore credits). Resize images to a maximum of 1024x1024 before converting to Base64 to save money.
• Use clear prompts: Tell the model exactly what to look for (e.g., "Extract the total amount from this receipt").
• Combine with Structured Output: You can ask the model to look at an image and return the data as a strict JSON object.