Built-in tools vs custom tools comparison
Quick answer
Use built-in tools for quick, standardized integrations with pre-defined capabilities, while custom tools offer flexibility to define specific functions and parameters tailored to your application. Both leverage the tools parameter in OpenAI chat completions but differ in control and complexity.
VERDICT
Use built-in tools for rapid deployment of common functionalities; choose custom tools when you need precise control over tool behavior and parameters.
| Tool type | Key strength | Pricing | API access | Best for |
|---|---|---|---|---|
| Built-in tools | Predefined, ready-to-use capabilities | Included in API usage | Direct via tools param | Standard tasks like calculator, browser |
| Custom tools | Fully customizable functions and parameters | Included in API usage | Defined via tools JSON schema | Unique app-specific integrations |
| Function calling (deprecated) | Simpler function calls with limited control | Included in API usage | Legacy functions param | Legacy codebases |
| Third-party plugins | Ecosystem integrations with external services | Varies by plugin | Via plugin marketplace | Extended capabilities beyond API |
Key differences
Built-in tools are predefined by the API provider with fixed capabilities, enabling quick integration without extra setup. Custom tools require you to define the tool schema, including name, description, and JSON parameters, giving you full control over the tool's behavior. Built-in tools simplify common tasks, while custom tools enable complex, domain-specific workflows.
Side-by-side example
Using a built-in calculator tool to perform a math operation:
from openai import OpenAI
import os
client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
# Built-in calculator tool usage
response = client.chat.completions.create(
model="gpt-4o-mini",
tools=[{"type": "calculator"}],
messages=[{"role": "user", "content": "Calculate 12 * 15"}]
)
print(response.choices[0].message.content) output
The result of 12 * 15 is 180.
Custom tool equivalent
Defining and using a custom tool for a weather query with explicit parameters:
from openai import OpenAI
import os
import json
client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
weather_tool = [{
"type": "function",
"function": {
"name": "get_weather",
"description": "Get weather for a location",
"parameters": {
"type": "object",
"properties": {
"location": {"type": "string"}
},
"required": ["location"]
}
}
}]
response = client.chat.completions.create(
model="gpt-4o-mini",
tools=weather_tool,
messages=[{"role": "user", "content": "What's the weather in NYC?"}]
)
if response.choices[0].finish_reason == "tool_calls":
tool_call = response.choices[0].message.tool_calls[0]
args = json.loads(tool_call.function.arguments)
print(f"Tool called: {tool_call.function.name} with args: {args}")
else:
print(response.choices[0].message.content) output
Tool called: get_weather with args: {'location': 'NYC'} When to use each
Use built-in tools when you need fast, reliable access to common functionalities without extra development. Opt for custom tools when your application requires specific inputs, outputs, or workflows not covered by built-in options. Custom tools are essential for integrating proprietary APIs or complex logic.
| Use case | Recommended tool type | Reason |
|---|---|---|
| Quick math or calculator | Built-in tools | Predefined and optimized for speed |
| Domain-specific API integration | Custom tools | Full control over parameters and behavior |
| Standardized browsing or search | Built-in tools | Ready-made and maintained by provider |
| Unique business logic or workflows | Custom tools | Tailored to exact application needs |
Pricing and access
| Option | Free | Paid | API access |
|---|---|---|---|
| Built-in tools | Yes (included) | Billed as normal API usage | Direct via tools param |
| Custom tools | Yes (included) | Billed as normal API usage | Defined via tools JSON schema |
| Function calling (deprecated) | Yes | Billed as normal API usage | Legacy functions param |
| Third-party plugins | Varies | Varies by plugin | Via plugin marketplace |
Key Takeaways
- Built-in tools provide fast, standardized capabilities with minimal setup.
- Custom tools offer full control for complex or unique application needs.
- Use the tools parameter in OpenAI chat completions for both types.
- Function calling is deprecated; migrate to tools for new integrations.
- Pricing for both built-in and custom tools is included in standard API usage.