Skip to main content

​
Gemini Tools

Gemini models support multiple built-in tools, including code execution, Google Search, and URL context retrieval. 
POST /v1beta/models/{model}:generateContent

​
Code Execution

Let Gemini write and execute Python code: 
curl "https://crazyrouter.com/v1beta/models/gemini-2.5-flash:generateContent?key=YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {
        "role": "user",
        "parts": [
          {"text": "Calculate the sum of the first 100 prime numbers"}
        ]
      }
    ],
    "tools": [
      {
        "codeExecution": {}
      }
    ]
  }'

​
Code Execution Response

{
  "candidates": [
    {
      "content": {
        "parts": [
          {
            "text": "I will calculate the sum of the first 100 prime numbers."
          },
          {
            "executableCode": {
              "language": "PYTHON",
              "code": "def is_prime(n):\n    if n < 2:\n        return False\n    for i in range(2, int(n**0.5) + 1):\n        if n % i == 0:\n            return False\n    return True\n\nprimes = []\nn = 2\nwhile len(primes) < 100:\n    if is_prime(n):\n        primes.append(n)\n    n += 1\n\nprint(f'Sum of first 100 primes: {sum(primes)}')"
            }
          },
          {
            "codeExecutionResult": {
              "outcome": "OUTCOME_OK",
              "output": "Sum of first 100 primes: 24133\n"
            }
          },
          {
            "text": "The sum of the first 100 prime numbers is **24133**."
          }
        ],
        "role": "model"
      }
    }
  ]
}
Enable Google Search to let the model access real-time information: 
curl "https://crazyrouter.com/v1beta/models/gemini-2.5-flash:generateContent?key=YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {
        "role": "user",
        "parts": [
          {"text": "What are the latest AI technology trends in 2026?"}
        ]
      }
    ],
    "tools": [
      {
        "googleSearch": {}
      }
    ]
  }'

​
URL Context

Let the model read the content of a specified URL:
cURL
curl "https://crazyrouter.com/v1beta/models/gemini-2.5-flash:generateContent?key=YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {
        "role": "user",
        "parts": [
          {"text": "Summarize the main content of this webpage: https://example.com/article"}
        ]
      }
    ],
    "tools": [
      {
        "urlContext": {}
      }
    ]
  }'

​
Custom Function Calling

Gemini also supports custom function calling (similar to OpenAI’s Function Calling):
cURL
curl "https://crazyrouter.com/v1beta/models/gemini-2.5-flash:generateContent?key=YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "contents": [
      {
        "role": "user",
        "parts": [{"text": "Check the weather in Beijing"}]
      }
    ],
    "tools": [
      {
        "functionDeclarations": [
          {
            "name": "get_weather",
            "description": "Get weather information for a specified city",
            "parameters": {
              "type": "object",
              "properties": {
                "city": {
                  "type": "string",
                  "description": "City name"
                }
              },
              "required": ["city"]
            }
          }
        ]
      }
    ]
  }'

​
Function Call Response

{
  "candidates": [
    {
      "content": {
        "parts": [
          {
            "functionCall": {
              "name": "get_weather",
              "args": {"city": "Beijing"}
            }
          }
        ]
      }
    }
  ]
}

​
Return Function Result

{
  "contents": [
    {"role": "user", "parts": [{"text": "Check the weather in Beijing"}]},
    {"role": "model", "parts": [{"functionCall": {"name": "get_weather", "args": {"city": "Beijing"}}}]},
    {
      "role": "user",
      "parts": [
        {
          "functionResponse": {
            "name": "get_weather",
            "response": {"temperature": 18, "condition": "sunny"}
          }
        }
      ]
    }
  ]
}
Multiple tools can be enabled simultaneously. For example, enabling both code execution and Google Search allows the model to automatically choose the appropriate tool as needed.