Do the answers come back with sources attached?

Yes. Create Chat Completion runs against Perplexity Sonar, which returns web-grounded responses with citations, so an answer can be shown alongside where it came from. Perplexity Search (Raw Results) goes further and returns ranked results straight from the index with no model summarization at all.

What is the difference between the sync and async completion actions?

Create Chat Completion returns in the same call and suits a live conversation. Create Async Chat Completion is for long-running work: it returns an ID that Get Async Chat Completion retrieves later, and List Async Chat Completions shows everything queued on the account.

Whose Perplexity usage does this consume?

The API key you connect. Every completion, agent run, search and embedding call bills to that account rather than to a shared pool, so a Perplexity plan that fits the volume you expect is worth confirming before putting it in front of live conversations.

Which model does a request use?

Whichever you specify. List Models returns the identifiers available for the Agent API, and both the Sonar completion actions and Execute Agent accept a model parameter, so a cheaper model can handle routine lookups while a stronger one is reserved for the async research jobs.

Perplexity MCP integration

Returns cited, web-grounded answers in conversation, queues longer research as async jobs, and generates embeddings for your own text.

Connect Perplexity See all integrations

9actions available

Three actions you can hand over today

Every action runs live through MCP. Nothing to build, nothing to maintain.

Create chat completion
Lyro sends a query to Perplexity's Sonar model and gets back an answer grounded in current web search results. A support agent gets a sourced, up-to-date answer to a tricky question without leaving the conversation to research it.
Execute agent
Lyro runs a Perplexity agent to handle a multi-step request and returns its response. A team can hand off a research-heavy question and get a complete answer back instead of chaining several searches together manually.
Get async chat completion
Lyro checks on an asynchronous Perplexity chat completion using its ID and retrieves the result once ready. An agent that submitted a longer query can pick up the answer later instead of waiting on it in real time.

See all 9 actions

How businesses use Perplexity + Lyro

Each card is one request a support team gets, and the Perplexity actions Lyro runs to close it.

Answer from the live web, with sources
Lyro sends the question to Sonar and gets back a web-grounded answer with citations, or pulls the raw ranked search results when the underlying links matter more than the summary.
Create Chat CompletionPerplexity Search (Raw Results)
Queue research that takes too long to wait for
For a question that needs real digging, Lyro submits an async completion, collects the result by ID once it finishes, and lists what is still outstanding.
Create Async Chat CompletionGet Async Chat CompletionList Async Chat Completions
Run a multi-step agentic lookup
When one search is not enough, Lyro triggers an agent run with built-in web search and URL fetching, choosing from the models the Agent API currently exposes.
Execute AgentList Models
Embed your own text for retrieval
Lyro generates vector embeddings for queries, sentences or documents, and document-aware contextualized embeddings when chunks need to carry the surrounding context.
Create EmbeddingsCreate Contextualized Embeddings

How it works

Get started in 3 steps

Connect once, then just ask. There is no workflow builder to learn and nothing to maintain — Lyro reads the Perplexity actions it has and picks the ones a request needs.

01
Connect Perplexity
Authorize the Perplexity account your team already uses — one consent screen, no API keys, no mapping tables. Lyro can only do what you granted that account, and you can disconnect it at any time.
02
Tell your agent what you need
Describe the job the way you would hand it to a teammate. Lyro maps it to the Perplexity actions that close it and chains as many as the request needs.
03
Watch it work
The agent runs the actions inside the conversation the customer is already in, so nobody copies data between tabs and your team can take over at any point.
Get started free

Perplexity · Lyro

Everything else about Perplexity

Setup, permissions, and the limits of what Lyro can do inside Perplexity.

Yes. Create Chat Completion runs against Perplexity Sonar, which returns web-grounded responses with citations, so an answer can be shown alongside where it came from. Perplexity Search (Raw Results) goes further and returns ranked results straight from the index with no model summarization at all.

Every action available in Perplexity

All 9 actions your agent can call on Perplexity, straight from the live MCP connection.

Create async chat completion
Create Async Chat Completion (POST /v1/async/sonar).
Create chat completion
Perplexity Sonar Chat Completions (POST /v1/sonar).
Create contextualized embeddings
Create Contextualized Embeddings (POST /v1/contextualizedembeddings).
Createembeddings
Generate vector embeddings for independent texts (queries, sentences, documents).
Execute agent
Create Agent Response (POST /v1/agent).
Get async chat completion
Get Async Chat Completion (GET /v1/async/sonar/{id}).
List async chat completions
List Async Chat Completions (GET /v1/async/sonar).
List models
List Models (GET /v1/models).
Perplexity search (raw results)
Search the Web (POST /search).