---
Title: "What Is Tool Calling in AI? Function Calling"
Url: "https://computer.io/glossary/tool-calling"
Topic: "Agentic AI"
Description: "Tool calling, or function calling, lets an AI model request actions from external tools and APIs, like searching data or updating a record. How it works."
---

# What is tool calling in AI?

Tool calling, also called function calling, is the ability of an AI model to request that an external tool or API be run, such as searching a database, creating a ticket or issuing a refund, by outputting a structured call that the application executes and returns results from.

## How it works

1. Tools are described to the model (name, inputs).
2. The model decides to call one and outputs arguments.
3. The app runs the tool.
4. Results go back to the model.
5. The model continues or answers.

## Example

`get_order_status(order_id="4412")` → "shipped" → "Your order shipped yesterday."

## Safety

Limit tools by permission, validate inputs and require approval for risky actions.

## FAQ

### Is tool calling the same as function calling?

Yes, the terms are used interchangeably.

### How does MCP relate to tool calling?

MCP standardizes how tools are exposed to models.

### Can tool calling make mistakes?

Yes, models can choose the wrong tool or arguments; validate and test.

## Related terms

- [Model Context Protocol (MCP)](https://computer.io/glossary/model-context-protocol)
- [AI agent](https://computer.io/glossary/ai-agent)
- [AI guardrails](https://computer.io/glossary/ai-guardrails)

## How Computer helps

- [How computer works](https://computer.io/how-computer-works)
- [Agent studio](https://devrev.ai/agent-studio)

## Further reading

- [Build enterprise ai agents](https://devrev.ai/blog/build-enterprise-ai-agents)
