Integration

LLMLayer integration

LLMLayer is a web data API for AI agents. Search the web (general, news, shopping, videos, images, scholar), get an LLM-written answer grounded in live search results (optionally as JSON matching a schema), scrape a page as markdown, HTML or a screenshot, extract structured data, a summary, an answer to a question, links or a brand profile from one page, list the URLs of a website, and read YouTube transcripts and the text of public PDFs. Every call is billed to the LLMLayer account per the vendor's pricing page (for example $0.002 per web search), and the request rate starts at 20 per minute, rising with the credits purchased. The two streaming routes (the Server-Sent Events answer stream and the multi-page crawl stream) are not exposed, so there is no multi-page crawl tool. Authentication is an LLMLayer API key sent as a Bearer token.

Connect LLMLayer How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in LLMLayer.Built from the 7 LLMLayer tools below, on the connection you add once.

Common LLMLayer use cases

summarizing a YouTube video
finding every blog post URL on a site
finding pages to scrape about a topic
reading a report linked from a web page
pulling product name and price from a product page
reading the full text of a page found by web_search
answering a question with current information from the web
extracting quotes from a talk
feeding a public PDF into an LLM node
choosing pages to scrape from a domain

What you can do with LLMLayer

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Each badge shows the default agent permission of an endpoint (allowed: an agent calls it without asking; ask first: you are asked before an agent calls it, when the agent or chat asks permission for sensitive actions; never: no agent may call it), which every user can change endpoint by endpoint in their credential settings.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Pdf text, 1 operation

get_pdf_contentGet pdf contentPOSTAgents: ask firstExtracts the text of a public PDF from its direct URL (no password-protected or private files). Costs $0.002 per call. A slow PDF fails with HTTP 504 (the reference names a 15-second timeout).

Web answer, 1 operation

answer_questionAnswer questionPOSTAgents: ask firstSearches the web and has an LLM write an answer grounded in the results, in one billed call: llmlayer-web costs $0.007 per search query (max_queries, default 1). Set answer_type json with json_schema for structured output.

Web page, 1 operation

scrape_pageScrape pagePOSTAgents: ask firstFetches one web page (not a PDF) and returns it as markdown, html and/or a base64 PNG screenshot. Billed $0.001 per format requested, plus $0.004 with advanced_proxy. For a PDF use get_pdf_content.

Web page data, 1 operation

extract_pageExtract pagePOSTAgents: ask firstRuns one or more modes on one web page: json (structured_data, needs json_schema), summary, qa (answer, needs query), links, brand. Costs add up per mode: $0.005 per AI mode, links $0.001, brand $0.002.

Web search result, 1 operation

web_searchWeb searchPOSTAgents: ask firstReturns raw web search results (no LLM) for a query: general, news, shopping, videos, images or scholar. Costs $0.002 per call. Result fields vary by search_type (the vendor's example result has title, link and snippet).

Website url, 1 operation

map_websiteMap websitePOSTAgents: ask firstLists the URLs found on a website, each with its page title, without fetching page content. Flat $0.002 per call. Narrow with search; limit caps the count (vendor default 5000, maximum 30000).

Youtube transcript, 1 operation

get_youtube_transcriptGet youtube transcriptPOSTAgents: ask firstReturns the transcript of a public YouTube video that has captions, with its title, description, channel, views, likes and date when available. Costs $0.003 per call.

Use LLMLayer from Claude Code, Cursor or Codex

Connect LLMLayer in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its LLMLayer operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect LLMLayer

Create your workspace
Sign up free. Nothing to install.
Connect LLMLayer once
Add your LLMLayer connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.

LLMLayer integration FAQ

What can you do with the LLMLayer integration?

LiveContext exposes 7 operations on LLMLayer, covering pdf text, web answer, web page and web page data. Common uses include summarizing a YouTube video, finding every blog post URL on a site and finding pages to scrape about a topic.