SERPdive
The SERPdive tool connects your Dify agents and workflows to a web search API built for AI: one call returns clean, answer-ready content extracted from live pages instead of a list of links to fetch.
Quality is measured, not asserted: on a public 1,000-question benchmark judged blind by an independent model, SERPdive won 60.7% of decided quality duels against Tavily's default search while returning 20.2% fewer tokens at the same speed. The methodology and per-question results are open, so anyone can replay the run: serpdive-benchmark.
Available Actions
| Action | Description | Best For |
|---|
| Search | Web search that returns extracted, answer-ready page content, best sources first | Real-time information, fact-checking, RAG, research agents |
Two retrieval depths:
- Mako (1 credit): fast, returns the fact-carrying sentences of each source.
- Moby (1.5 credits): returns the full readable content of every page, for when whole-page context is needed.
Optionally, the API also writes an answer built from the sources (answer: true, no extra credits): concise with Mako, detailed with numbered citations with Moby.
Configuration
1. Get Your API Key
Sign up at serpdive.com and create a key in the dashboard.
Every account gets 1,000 free credits every month, no card required.
2. Install from Plugin Marketplace
Navigate to Plugin Marketplace in Dify and search for "SERPdive". Click Install.
3. Add Your API Key
Go to Tools → SERPdive → Authentication and paste your API key. Validating the key never spends credits.
4. Add a Tool node to your flow
Add a Node → Tool → SERPdive Search, or enable it as a tool in an Agent application.
Common Patterns
Fact-checking workflow:
User question → SERPdive Search (answer: true) → LLM validates and responds
Research agent:
Topic → SERPdive Search (model: moby) → LLM synthesizes from full-page content
Lean RAG:
Query → SERPdive Search (max_results: 3) → LLM answers from the extracted content, with a smaller token bill
Resources