Search API
AI-tuned relevance ranking adapts to entity, freshness, how-to, comparison, and multilingual research needs.
Search, extract, and retrieve trusted web sources with one simple API. SRCX turns web pages into clean, LLM-ready context for AI apps, agents, and RAG systems.
Traditional search APIs return links. SRCX handles the difficult layer after search: crawling, cleaning, deduplication, ranking, and context packaging for AI products.
AI-tuned relevance ranking adapts to entity, freshness, how-to, comparison, and multilingual research needs.
A resilient extraction layer removes navigation, ads, boilerplate, and duplicate text while preserving usable source context.
Structured context blocks combine trusted sources, concise excerpts, metadata, and citations for downstream AI workflows.
Every new account starts with one-time developer credits. Paid plans refresh monthly. Add-on credits are available on top of active membership.
One-time starter credits
Personal developers
Mid-size developers and small teams
Enterprise workloads
Start with cURL, then move into your backend. Results use a unified JSON envelope with source attribution, request IDs, and credit metadata.
GET /v1/search?q=Feishu%20what%20is&limit=5
Authorization: Bearer lmsk_...
{
"ok": true,
"data": {
"query_type": "entity_research",
"results": [
{"title":"Feishu", "source_type":"reference"}
]
}
}
Use SRCX anywhere an AI app needs web search, page extraction, citations, or fresh context. It works especially well for local AI tools, agent builders, RAG workflows, and GitHub open-source projects that need a clean search API.
Hermes Agent
Cherry Studio
OpenClaw
Open WebUI
Dify
LangChain
LM StudioAdd web awareness to desktop model tools and local assistants without building your own crawler, cleaner, cache, and ranking stack.
Let agents retrieve current sources, read clean pages, and pass cited context into prompts instead of guessing from stale model memory.
Connect SRCX through HTTP tools, custom actions, workflow nodes, or retrieval plugins in projects that need web search and extraction.
AI applications do not lack models. They lack reliable, timely, clean external information sources. SRCX is the source layer for agents, RAG systems, local AI tools, and developer products.