Perplexity AI logo
ConnectorPerplexity AI

Perplexity AI integration for Claude and Codex

Create a Perplexity AI connection, then control which Spaces can use its approved tools and data without pasting credentials into prompts or threads.

Use Perplexity AI from Claude

Add the Perplexity AI connection to a Space that runs Claude, then use it from every channel in that Space.

Use Perplexity AI from Codex

Add the Perplexity AI connection to a Space that runs Codex, then use it from every channel in that Space.

Use Perplexity AI from Sidekick

Connect your own OpenClaw to Sidekick while type.com provides the conversations, permissions, connections, skills, and automations.

Analytics & Data

What the Perplexity AI connector can do

Perplexity AI offers Sonar models that answer questions using live web search and return sources alongside the text, plus agent and embedding APIs. Add it as a connection in type.com and Claude can hand research questions to Perplexity from a shared thread, so your team gets cited findings, longer background research jobs, and vectors for search projects in one place.

One connection, many Spaces

Connect Perplexity AI once, then decide which Spaces can use it in threads, skills, automations, and coding work.

What teams do with Perplexity AI in type.com

Cited research inside a thread

Ask a market or product question and get an answer grounded in current web results, with source links the whole team can check before acting on it.

Try asking
Use Perplexity to find what changed in EU AI Act enforcement timelines this quarter, and list the sources it cites.

Long research jobs that run asynchronously

Submit a heavier research request as an async job so the agent can come back with a fuller write-up instead of a rushed single reply.

Try asking
Start an async Perplexity request comparing the five largest open-source vector databases on licensing and hosting, then summarize it here when done.

Embeddings for internal search

Turn product docs or support articles into vectors, including chunk-aware embeddings that keep document context, so engineering can prototype semantic search.

Try asking
Generate contextualized Perplexity embeddings for the chunks in our onboarding guide and save the output for the search prototype.

Representative actions

  • Create Async Chat Completion

    Queues a Sonar request that runs in the background, suited to research questions that take longer than a normal reply.

  • Create Chat Completion

    Returns a conversational answer based on live web search, along with the citations behind it.

  • Create Contextualized Embeddings

    Builds embeddings for document chunks that account for the surrounding document, useful for retrieval over long files.

  • CreateEmbeddings

    Produces vectors for standalone pieces of text such as queries or sentences, for similarity search and clustering.

  • Execute Agent

    Runs a Perplexity agent that can chain web search and page fetching across several steps to answer a complex request.

Connection

API and auth details

Perplexity exposes first-party Sonar, Search, Agent, and Embeddings APIs over HTTPS using bearer API tokens.

FAQ

Questions people ask before connecting Perplexity AI

Can Claude use Perplexity AI for web research in type.com?

Yes. When a Perplexity AI connection is shared with a Space, Claude can send questions to Sonar, start async research jobs, or run a Perplexity agent from any thread, and the cited answers stay visible to everyone in that thread.

How does Perplexity AI authentication work?

Perplexity uses a bearer API token. An admin adds the token once to the type.com connection and picks which Spaces may use it, so no one pastes the token into prompts or threads.

Can I schedule recurring Perplexity research?

Yes. Save a research prompt as a type.com automation, for example a Monday scan of competitor news, and the agent runs it with Perplexity on that schedule and posts the cited results to a channel.

Is this the same as a Perplexity AI MCP server?

Not exactly. type.com uses connectors and connections to give selected Spaces access to approved app tools and data. Some connectors use hosted MCP, while others use OAuth, API keys, service accounts, or custom APIs.

More analytics & data connectors

All Analytics & Data integrations