docs: Add comprehensive vision/multimodal support documentation

markdown/api-reference.mdx

@@ -409,6 +409,76 @@ curl -X POST http://localhost:8080/v1/chat/completions \
   }'
 ```
 
+### Vision/Multimodal Support
+
+For vision-capable models, you can include images in your requests using either HTTP URLs or base64-encoded data URLs. Vision support must be enabled with `ENABLE_VISION=true` in your configuration.
+
+#### Using HTTP URL
+
+```bash
+curl -X POST http://localhost:8080/v1/chat/completions \
+  -H "Content-Type: application/json" \
+  -d '{
+    "model": "anthropic/claude-3-5-sonnet-20241022",
+    "messages": [
+      {
+        "role": "user",
+        "content": [
+          {
+            "type": "text",
+            "text": "What is in this image?"
+          },
+          {
+            "type": "image_url",
+            "image_url": {
+              "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
+            }
+          }
+        ]
+      }
+    ]
+  }'
+```
+
+#### Using Base64 Data URL
+
+```bash
+curl -X POST http://localhost:8080/v1/chat/completions \
+  -H "Content-Type: application/json" \
+  -d '{
+    "model": "anthropic/claude-3-5-sonnet-20241022",
+    "messages": [
+      {
+        "role": "user",
+        "content": [
+          {
+            "type": "text",
+            "text": "Describe this image"
+          },
+          {
+            "type": "image_url",
+            "image_url": {
+              "url": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mP8z8DwHwAFBQIAX8jx0gAAAABJRU5ErkJggg=="
+            }
+          }
+        ]
+      }
+    ]
+  }'
+```
+
+**Supported Providers with Vision:**
+
+- OpenAI (GPT-4o, GPT-5, GPT-4.1, GPT-4 Turbo)
+- Anthropic (Claude 3, Claude 4, Claude 4.5 Sonnet, Claude 4.5 Haiku)
+- Google (Gemini 2.5)
+- Cohere (Command A Vision, Aya Vision)
+- Ollama (LLaVA, Llama 4, Llama 3.2 Vision)
+- Groq (vision models)
+- Mistral (Pixtral)
+
+**Note:** When `ENABLE_VISION=false` (default), requests containing image content will be rejected even if the model supports vision. This is disabled by default for performance and security reasons.
+
 ### Direct API Proxy
 
 For more advanced use cases, you can proxy requests directly to the provider's API:

markdown/configuration.mdx

@@ -21,6 +21,11 @@ Environment variables are the primary method for configuring Inference Gateway.
 <ConfigTable
   rows={[
     { variable: 'ENVIRONMENT', description: 'Deployment environment', defaultValue: 'production' },
+    {
+      variable: 'ENABLE_VISION',
+      description: 'Enable vision/multimodal support for all providers',
+      defaultValue: 'false',
+    },
     {
       variable: 'ENABLE_TELEMETRY',
       description: 'Enable OpenTelemetry metrics and tracing',
@@ -30,6 +35,8 @@ Environment variables are the primary method for configuring Inference Gateway.
   ]}
 />
 
+When `ENABLE_VISION` is set to `true`, Inference Gateway enables vision/multimodal capabilities, allowing you to send images alongside text in chat completion requests. When disabled (default), requests with image content will be rejected even if the provider and model support vision. This is disabled by default for performance and security reasons.
+
 When `ENABLE_TELEMETRY` is set to `true`, Inference Gateway exposes a `/metrics` endpoint for Prometheus scraping and generates distributed traces that can be collected by OpenTelemetry collectors.
 
 ### OpenID Connect
@@ -369,6 +376,7 @@ Here's a comprehensive example for configuring Inference Gateway in a production
 ```bash
 # General settings
 ENVIRONMENT=production
+ENABLE_VISION=true
 ENABLE_TELEMETRY=true
 ENABLE_AUTH=true
 

markdown/examples.mdx

@@ -98,4 +98,75 @@ curl -X POST http://localhost:8080/v1/chat/completions \
   }'
 ```
 
+### Vision/Multimodal Image Processing
+
+Process images with vision-capable models. First, enable vision support:
+
+```bash
+ENABLE_VISION=true
+```
+
+#### Using HTTP Image URL
+
+```bash
+curl -X POST http://localhost:8080/v1/chat/completions \
+  -H "Content-Type: application/json" \
+  -d '{
+    "model": "anthropic/claude-3-5-sonnet-20241022",
+    "messages": [
+      {
+        "role": "user",
+        "content": [
+          {
+            "type": "text",
+            "text": "What is in this image?"
+          },
+          {
+            "type": "image_url",
+            "image_url": {
+              "url": "https://upload.wikimedia.org/wikipedia/commons/thumb/d/dd/Gfp-wisconsin-madison-the-nature-boardwalk.jpg/2560px-Gfp-wisconsin-madison-the-nature-boardwalk.jpg"
+            }
+          }
+        ]
+      }
+    ]
+  }'
+```
+
+#### Using Base64 Data URL
+
+```bash
+curl -X POST http://localhost:8080/v1/chat/completions \
+  -H "Content-Type: application/json" \
+  -d '{
+    "model": "anthropic/claude-3-5-sonnet-20241022",
+    "messages": [
+      {
+        "role": "user",
+        "content": [
+          {
+            "type": "text",
+            "text": "What color is this pixel?"
+          },
+          {
+            "type": "image_url",
+            "image_url": {
+              "url": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAAEAAAABCAYAAAAfFcSJAAAADUlEQVR42mP8z8DwHwAFBQIAX8jx0gAAAABJRU5ErkJggg=="
+            }
+          }
+        ]
+      }
+    ]
+  }'
+```
+
+**Supported Vision Models:**
+
+- `anthropic/claude-3-5-sonnet-20241022` (Claude 4.5 Sonnet)
+- `anthropic/claude-3-5-haiku-20250219` (Claude 4.5 Haiku)
+- `openai/gpt-4o`
+- `google/gemini-2.5-flash`
+- `ollama/llava`
+- And more...
+
 For more detailed examples and use cases, check out the full [examples directory](https://github.com/inference-gateway/inference-gateway/tree/main/examples) in the GitHub repository.

markdown/supported-providers.mdx

@@ -11,61 +11,153 @@ The following LLM providers are currently supported:
   <div className="border p-4 rounded-md">
     <h3>OpenAI</h3>
     <p>Access GPT models including GPT-3.5, GPT-4, and more.</p>
-    <p><strong>Authentication:</strong> Bearer Token</p>
-    <p><strong>Default URL:</strong> https://api.openai.com/v1</p>
+    <div><strong>Authentication:</strong> Bearer Token</div>
+    <div><strong>Default URL:</strong> https://api.openai.com/v1</div>
+    <div><strong>Vision Support:</strong> ✅ Yes (GPT-4o, GPT-5, GPT-4.1, GPT-4 Turbo)</div>
   </div>
 
-  <div className="border p-4 rounded-md">
-    <h3>DeepSeek</h3>
-    <p>Use DeepSeek's models for various natural language tasks.</p>
-    <p><strong>Authentication:</strong> Bearer Token</p>
-    <p><strong>Default URL:</strong> https://api.deepseek.com</p>
+<div className="border p-4 rounded-md">
+  <h3>DeepSeek</h3>
+  <p>Use DeepSeek's models for various natural language tasks.</p>
+  <div>
+    <strong>Authentication:</strong> Bearer Token
   </div>
-
-  <div className="border p-4 rounded-md">
-    <h3>Anthropic</h3>
-    <p>Connect to Claude models for high-quality conversational AI.</p>
-    <p><strong>Authentication:</strong> X-Header</p>
-    <p><strong>Default URL:</strong> https://api.anthropic.com/v1</p>
+  <div>
+    <strong>Default URL:</strong> https://api.deepseek.com
   </div>
-
-  <div className="border p-4 rounded-md">
-    <h3>Cohere</h3>
-    <p>Use Cohere's models for various natural language tasks.</p>
-    <p><strong>Authentication:</strong> Bearer Token</p>
-    <p><strong>Default URL:</strong> https://api.cohere.com</p>
+  <div>
+    <strong>Vision Support:</strong> ❌ No
   </div>
-
-  <div className="border p-4 rounded-md">
-    <h3>Groq</h3>
-    <p>Access high-performance inference with Groq's LPU-accelerated models.</p>
-    <p><strong>Authentication:</strong> Bearer Token</p>
-    <p><strong>Default URL:</strong> https://api.groq.com/openai/v1</p>
+</div>
+
+<div className="border p-4 rounded-md">
+  <h3>Anthropic</h3>
+  <p>Connect to Claude models for high-quality conversational AI.</p>
+  <div>
+    <strong>Authentication:</strong> X-Header
   </div>
-
-  <div className="border p-4 rounded-md">
-    <h3>Cloudflare</h3>
-    <p>Connect to Cloudflare Workers AI for inference on various models.</p>
-    <p><strong>Authentication:</strong> Bearer Token</p>
-    <p><strong>Default URL:</strong> https://api.cloudflare.com/client/v4/accounts/</p>
-    <p className="text-sm">{'{ACCOUNT_ID}'}/ai</p>
+  <div>
+    <strong>Default URL:</strong> https://api.anthropic.com/v1
   </div>
-
-  <div className="border p-4 rounded-md">
-    <h3>Ollama</h3>
-    <p>Run open-source models locally or on a self-hosted server.</p>
-    <p><strong>Authentication:</strong> None (optional API key)</p>
-    <p><strong>Default URL:</strong> http://ollama:8080/v1</p>
+  <div>
+    <strong>Vision Support:</strong> ✅ Yes (Claude 3, Claude 4, Claude 4.5 Sonnet, Claude 4.5
+    Haiku)
+  </div>
+</div>
+
+<div className="border p-4 rounded-md">
+  <h3>Cohere</h3>
+  <p>Use Cohere's models for various natural language tasks.</p>
+  <div>
+    <strong>Authentication:</strong> Bearer Token
+  </div>
+  <div>
+    <strong>Default URL:</strong> https://api.cohere.com
+  </div>
+  <div>
+    <strong>Vision Support:</strong> ✅ Yes (Command A Vision, Aya Vision)
+  </div>
+</div>
+
+<div className="border p-4 rounded-md">
+  <h3>Groq</h3>
+  <p>Access high-performance inference with Groq's LPU-accelerated models.</p>
+  <div>
+    <strong>Authentication:</strong> Bearer Token
+  </div>
+  <div>
+    <strong>Default URL:</strong> https://api.groq.com/openai/v1
+  </div>
+  <div>
+    <strong>Vision Support:</strong> ✅ Yes (vision models)
+  </div>
+</div>
+
+<div className="border p-4 rounded-md">
+  <h3>Cloudflare</h3>
+  <p>Connect to Cloudflare Workers AI for inference on various models.</p>
+  <div>
+    <strong>Authentication:</strong> Bearer Token
+  </div>
+  <div>
+    <strong>Default URL:</strong> https://api.cloudflare.com/client/v4/accounts/
+  </div>
+  <div className="text-sm">{'{ACCOUNT_ID}'}/ai</div>
+  <div>
+    <strong>Vision Support:</strong> ❌ No
+  </div>
+</div>
+
+<div className="border p-4 rounded-md">
+  <h3>Ollama</h3>
+  <p>Run open-source models locally or on a self-hosted server.</p>
+  <div>
+    <strong>Authentication:</strong> None (optional API key)
   </div>
-
+  <div>
+    <strong>Default URL:</strong> http://ollama:8080/v1
+  </div>
+  <div>
+    <strong>Vision Support:</strong> ✅ Yes (LLaVA, Llama 4, Llama 3.2 Vision)
+  </div>
+</div>
+
   <div className="border p-4 rounded-md">
     <h3>Google</h3>
     <p>Access Google's Gemini models for text generation and understanding.</p>
-    <p><strong>Authentication:</strong> Bearer Token</p>
-    <p><strong>Default URL:</strong> https://generativelanguage.googleapis.com/v1</p>
+    <div><strong>Authentication:</strong> Bearer Token</div>
+    <div><strong>Default URL:</strong> https://generativelanguage.googleapis.com/v1</div>
+    <div><strong>Vision Support:</strong> ✅ Yes (Gemini 2.5)</div>
   </div>
 </div>
 
+## Vision/Multimodal Support
+
+Several providers support vision/multimodal capabilities, allowing you to process images alongside text. To use vision features, you must enable them in your configuration:
+
+```bash
+ENABLE_VISION=true
+```
+
+**Note:** Vision support is disabled by default for performance and security reasons. When disabled, requests containing image content will be rejected even if the model supports vision.
+
+### Providers with Vision Support
+
+- **OpenAI**: GPT-4o, GPT-5, GPT-4.1, GPT-4 Turbo
+- **Anthropic**: Claude 3, Claude 4, Claude 4.5 Sonnet, Claude 4.5 Haiku
+- **Google**: Gemini 2.5
+- **Cohere**: Command A Vision, Aya Vision
+- **Ollama**: LLaVA, Llama 4, Llama 3.2 Vision
+- **Groq**: Vision models
+- **Mistral**: Pixtral
+
+### Example Vision Request
+
+```bash
+curl -X POST http://localhost:8080/v1/chat/completions \
+  -H "Content-Type: application/json" \
+  -d '{
+    "model": "anthropic/claude-3-5-sonnet-20241022",
+    "messages": [
+      {
+        "role": "user",
+        "content": [
+          {
+            "type": "text",
+            "text": "What is in this image?"
+          },
+          {
+            "type": "image_url",
+            "image_url": {
+              "url": "https://example.com/image.jpg"
+            }
+          }
+        ]
+      }
+    ]
+  }'
+```
+
 ## Using Providers
 
 ### Provider Configuration