---
title: Image Generation
description: Generate images using AI models through the Agnic AI Gateway
---

# Image Generation

Generate images from text prompts using AI models with image output capabilities. Agnic supports various image generation models that can create high-quality images based on your descriptions.

## Overview

Image generation is available through models that have `"image"` in their `output_modalities`. These models can create images from text prompts when you specify the appropriate modalities in your request.

---

## Basic Image Generation

### Python

```python
from openai import OpenAI

client = OpenAI(
    api_key="agnic_tok_YOUR_TOKEN",
    base_url="https://api.agnic.ai/v1"
)

response = client.chat.completions.create(
    model="google/gemini-2.0-flash",
    messages=[
        {
            "role": "user",
            "content": "Generate a beautiful sunset over mountains"
        }
    ],
    extra_body={
        "modalities": ["image", "text"]
    }
)

# The generated image will be in the response
message = response.choices[0].message
if hasattr(message, 'images') and message.images:
    for image in message.images:
        image_url = image["image_url"]["url"]  # Base64 data URL
        print(f"Generated image: {image_url[:50]}...")
```

### JavaScript

```javascript
import OpenAI from 'openai';

const client = new OpenAI({
  apiKey: 'agnic_tok_YOUR_TOKEN',
  baseURL: 'https://api.agnic.ai/v1'
});

const response = await client.chat.completions.create({
  model: 'google/gemini-2.0-flash',
  messages: [
    {
      role: 'user',
      content: 'Generate a beautiful sunset over mountains'
    }
  ],
  modalities: ['image', 'text']
});

// The generated image will be in the response
const message = response.choices[0].message;
if (message.images) {
  message.images.forEach((image, index) => {
    const imageUrl = image.image_url.url; // Base64 data URL
    console.log(`Generated image ${index + 1}: ${imageUrl.substring(0, 50)}...`);
  });
}
```

### cURL

```bash
curl https://api.agnic.ai/v1/chat/completions \
  -H "Authorization: Bearer agnic_tok_YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "google/gemini-2.0-flash",
    "messages": [
      {
        "role": "user",
        "content": "Generate a beautiful sunset over mountains"
      }
    ],
    "modalities": ["image", "text"]
  }'
```

---

## Image Configuration Options

Some models support additional configuration through the `image_config` parameter.

### Aspect Ratio

Set `image_config.aspect_ratio` to request specific aspect ratios:

| Ratio | Resolution | Use Case |
|-------|------------|----------|
| `1:1` | 1024×1024 | Social media, avatars |
| `16:9` | 1344×768 | Widescreen, presentations |
| `9:16` | 768×1344 | Mobile, stories |
| `4:3` | 1184×864 | Standard photos |
| `3:2` | 1248×832 | Classic photography |

### Example with Configuration

```python
response = client.chat.completions.create(
    model="google/gemini-2.0-flash",
    messages=[
        {
            "role": "user",
            "content": "Create a panoramic landscape of a futuristic city"
        }
    ],
    extra_body={
        "modalities": ["image", "text"],
        "image_config": {
            "aspect_ratio": "16:9"
        }
    }
)
```

---

## Response Format

Generated images are returned in the assistant message:

```json
{
  "choices": [
    {
      "message": {
        "role": "assistant",
        "content": "I've generated a beautiful sunset image for you.",
        "images": [
          {
            "type": "image_url",
            "image_url": {
              "url": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAA..."
            }
          }
        ]
      }
    }
  ]
}
```

### Image Format Details

- **Format**: Images are returned as base64-encoded data URLs
- **Type**: Typically PNG format (`data:image/png;base64,`)
- **Multiple Images**: Some models can generate multiple images
- **Size**: Dimensions vary by model and configuration

---

## Saving Generated Images

### Python

```python
import base64

# Extract base64 data from response
image_data = response.choices[0].message.images[0]["image_url"]["url"]
# Remove the data URL prefix
base64_data = image_data.split(",")[1]

# Decode and save
with open("generated_image.png", "wb") as f:
    f.write(base64.b64decode(base64_data))
```

### JavaScript

```javascript
import fs from 'fs';

const imageData = response.choices[0].message.images[0].image_url.url;
const base64Data = imageData.split(',')[1];

fs.writeFileSync('generated_image.png', Buffer.from(base64Data, 'base64'));
```

---

## Compatible Models

Models with image generation capabilities:

| Model | Provider | Features |
|-------|----------|----------|
| `google/gemini-2.0-flash` | Google | Fast generation, configurable |

<Callout type="info">
  Check the model's `architecture.output_modalities` for `"image"` support. Visit [AI Gateway Pricing](https://app.agnic.ai/ai-gateway/pricing) for the full list.
</Callout>

---

## Best Practices

1. **Clear prompts** - Provide detailed descriptions for better results
2. **Specify style** - Include artistic style, mood, lighting preferences
3. **Set aspect ratio** - Match the ratio to your use case
4. **Error handling** - Check for the `images` field before processing

---

## Troubleshooting

**No images in response?**
- Verify the model supports image generation
- Ensure `modalities: ["image", "text"]` is in your request
- Check that your prompt requests image generation

**Wrong dimensions?**
- Use the `image_config.aspect_ratio` parameter
- Check model-specific documentation for supported ratios

<Cards>
  <Card title="Image Inputs" href="/docs/ai-gateway/multimodal/images" />
  <Card title="Video Inputs" href="/docs/ai-gateway/multimodal/video" />
</Cards>