Skip to main content

​
Web Search

Some models support performing web searches during conversations to retrieve real-time information before generating responses. Enable this by adding a web_search tool in tools.

​
Basic Usage

curl https://crazyrouter.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "gpt-4o",
    "messages": [
      {"role": "user", "content": "What are today'\''s top tech news stories?"}
    ],
    "tools": [
      {
        "type": "web_search"
      }
    ]
  }'
Python
stream = client.chat.completions.create(
    model="gpt-4o",
    messages=[
        {"role": "user", "content": "What are the latest AI models in 2026?"}
    ],
    tools=[{"type": "web_search"}],
    stream=True
)

for chunk in stream:
    if chunk.choices[0].delta.content is not None:
        print(chunk.choices[0].delta.content, end="")

​
Search Configuration

You can control search behavior through the web_search tool parameters: 
{
  "tools": [
    {
      "type": "web_search",
      "web_search": {
        "enable": true,
        "search_context_size": "medium"
      }
    }
  ]
}
ParameterTypeDescription
enablebooleanWhether to enable search
search_context_sizestringSearch context size: low, medium, high
Web search increases response latency and token consumption. The model automatically determines whether a search is needed and only triggers a search when real-time information is required.
Not all models support web search. If the model does not support it, the web_search tool will be ignored.