Back to APIs
Perplexity AI
Perplexity's Sonar models combine real-time web search with LLM reasoning — get grounded, cited answers using an OpenAI-compatible API.
Base URL
https://api.perplexity.aisearchRAGcitationsreal-timeSonar
Endpoints
POST
/chat/completionsGet a grounded answer with real-time web citations.
const response = await fetch("https://api.perplexity.ai/chat/completions", {
method: "POST",
headers: {
"Content-Type": "application/json",
"Authorization": "Bearer YOUR_ACCESS_TOKEN"
},
body: JSON.stringify({
"model": "sonar",
"messages": [
{
"role": "user",
"content": "What happened in tech news today?"
}
]
})
});
const data = await response.json();
console.log(data);Response Preview
{
"choices": [{
"message": {
"role": "assistant",
"content": "Here are today's top tech stories..."
}
}],
"citations": ["https://techcrunch.com/...", "https://theverge.com/..."]
}