Skip to main content
We’re excited to introduce Fast, a new depth parameter for the Linkup /search endpoint. Fast is our sub-second search mode, optimized to provide accurate results and snippets in the lowest possible time. Fast works best with focused queries, where you’re looking for one specific piece of information.

What’s New

fast uses a simpler, more direct search pipeline than standard or deep. The key difference is scope:
  • fast — one piece of information, one direct answer. Best for short, focused prompts. Only search results and relevant content snippets are returned.
  • standard — best to handle several topics, data sources (general web search, financial data, LinkedIn public data), or retrieve fully scraped content of a URL provided in the query.
This makes fast ideal for conversational use cases, real-time lookups, and high-volume pipelines where latency is critical.
The fast depth is currently in beta. We welcome any feedback — reach out on Discord or at support@linkup.so.

How to Use

A simple rule of thumb: if your prompt is short and you’re looking for one specific thing, use fast. If your prompt is longer or spans multiple topics, use standard.
Use fastUse standard
”Who is the CEO of OpenAI?""Find the pricing, features, and customer reviews for Notion"
"Current EUR/USD exchange rate""Current EUR/USD exchange rate and analysts analysis"
"What is Linkup’s website?""What is Linkup’s website and LinkedIn URL”
Simply set depth="fast" in your API request:
curl --request POST \
  --url https://api.linkup.so/v1/search \
  --header 'Authorization: Bearer {{LINKUP_API_KEY}}' \
  --header 'Content-Type: application/json' \
  --data '{
  "q": "weather in Paris today",
  "depth": "fast",
  "outputType": "sourcedAnswer"
}'

When to Use Each Depth

DepthBest For
fast (beta)Simple, focused queries — one piece of information, low latency
standardGeneral queries, multiple topics or sources
deepSequential queries — search for info, then scrape or search again based on results

Pricing

Fast depth costs €0.005 per call, the same as Standard. For more details, see Concepts.