Agent Fast APIPricing, Parameters & Features
Learn how to use Agent Fast. See the API price, context window, parameters, capabilities, and examples. Test the model directly in AI API Console.
fastWhat is Agent Fast?
Intended use: fast answers and lightweight web research. (Legacy equivalent: sonar)
You can send the model a prompt and receive a response. Depending on the model features, you can also send images, use tools, and ask for structured data.
AI API Console brings the important model information into one place, so you can understand the API before you start building.
Everything in one place
Check pricing, limits, parameters, features, and API examples without jumping between pages.
Agent Fast specifications
The main technical details developers should know before using the Agent Fast API.
The company that provides this model.
The identifier used in API requests.
The amount of information the model can work with.
Types of input supported by this model.
The main type of response generated.
Use the model from your own application.
What can Agent Fast do?
These are common features developers can use when working with the model API.
Text Generation
Create answers, summaries, explanations, ideas, and other text.
Structured Output
Return information in a predictable structured format such as JSON.
Web Search
Connect the model with live web results.
Streaming
Receive the response as it is generated instead of waiting for everything.
API Testing
Test prompts and settings before adding the model to your application.
Agent Fast API pricing
API cost is based on tokens. Input tokens are the information you send to the model. Output tokens are the information the model generates.
Tokens sent to the model.
Tokens generated by the model.
Your final API bill depends on how many tokens your application uses. Always check the provider's latest pricing before using a model in production.
Agent Fast API parameters
Parameters let you control how the model creates its response. Supported parameters can change between models.
temperatureControls how varied the response can be.
top_pControls how many possible next tokens are considered.
max_output_tokensSets the maximum size of the generated response.
streamReturns the response while it is being generated.
web_searchLets the model search the web for recent information.
Agent Fast API example
Here is a simple example of an API request. You can use the same idea in your own application.
curl https://api.perplexity.ai/v1/agent \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $API_KEY" \
-d '{
"model": "fast",
"messages": [
{
"role": "user",
"content": "Explain artificial intelligence simply."
}
]
}'Agent Fast API questions
What is Agent Fast?
Agent Fast is an AI model from Perplexity that developers can use through an API to build AI-powered applications.
How much does Agent Fast cost?
Agent Fast uses token-based API pricing. Input and output tokens have different prices.
What is the Agent Fast context window?
Agent Fast has a context window of 127,000 tokens based on the current model configuration.
What is the Agent Fast model ID?
The model ID used by this page is fast.
How can I test Agent Fast?
You can use the AI API Console playground to test prompts, parameters, and API responses.
Does Agent Fast support images?
No, Agent Fast does not currently support image input according to the capabilities configured for this model.