> For a complete page index, fetch https://docs.synthflow.ai/llms.txt. For full documentation content, fetch https://docs.synthflow.ai/llms-full.txt.

# Dynamic prompt injection

> Dynamically change an agent prompt at runtime so call behavior can adapt to external context and API inputs.

> **Note**
>
> Customize the behavior of your Voice AI Phone Call Agents by dynamically changing the prompt the model uses.

## Overview

Dynamic Prompt Injection allows you to tailor the interactions of your AI agents, providing a more personalized and contextually relevant experience. By sending customized prompts via API requests, you can guide the AI's conversation flow based on specific user needs or situational requirements.

## Use Cases

* **Customized Interactions**: Adapt the AI agent's behavior to suit different scenarios, such as customer support, sales inquiries, or technical assistance.
* **Contextual Responses**: Generate responses that are relevant to the current conversation, taking into account the user's previous inputs or questions.
* **Per-conversation Adaptability**: Change the AI's approach or tone, depending on the conversation so far or the interaction the customer has had.

## How to Use Dynamic Prompt Injection

* **Integrate Prompts into API Requests**: Send these custom prompts as part of your API request payload. This enables the AI to adapt its responses based on the provided prompt.

```json
{
 "prompt": "You are a customer service agent assisting with product inquiries."
}
```