Context-Aware Semantic Retrieval
Searches unstructured documentation by intent, technical concept, and context rather than exact keyword string matching.
Loading RSFlowHub...
Semantic knowledge base search API delivering context-aware document chunk retrieval for RAG workflows.
https://test.rsflowhub.com/api/v1/ai/knowledge-bases
Semantic Knowledge Base Vector Search API is a ready-to-use REST API on RSFlowHub for developers building web, mobile and AI agent applications. Send a JSON request to POST /api/v1/knowledge-bases with your API key; RSFlowHub runs it on a tuned AI model (with automatic fallback to a backup provider) and returns a structured JSON result, usually within 1 to 3 seconds.
X-API-Key Header
curl -X POST "https://test.rsflowhub.com/api/v1/knowledge-bases" \
-H "Content-Type: application/json" \
-H "X-API-Key: rfh_your_api_key" \
-d '{"name":"Customer Support KB","description":"Knowledge base containing product FAQs, return policies, and documentation."}'
<?php
$client = new \GuzzleHttp\Client();
$response = $client->post('https://test.rsflowhub.com/api/v1/knowledge-bases', [
'headers' => [
'Content-Type' => 'application/json',
'X-API-Key' => 'rfh_your_api_key',
],
'json' => array (
'name' => 'Customer Support KB',
'description' => 'Knowledge base containing product FAQs, return policies, and documentation.',
),
]);
echo $response->getBody();
const response = await fetch('https://test.rsflowhub.com/api/v1/knowledge-bases', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
'X-API-Key': 'rfh_your_api_key'
},
body: JSON.stringify({"name":"Customer Support KB","description":"Knowledge base containing product FAQs, return policies, and documentation."})
});
const data = await response.json();
console.log(data);
import requests
url = "https://test.rsflowhub.com/api/v1/knowledge-bases"
headers = {
"Content-Type": "application/json",
"X-API-Key": "rfh_your_api_key"
}
payload = {"name":"Customer Support KB","description":"Knowledge base containing product FAQs, return policies, and documentation."}
response = requests.post(url, json=payload, headers=headers)
print(response.json())
Searches unstructured documentation by intent, technical concept, and context rather than exact keyword string matching.
Combines lexical BM25 keyword matching with dense vector embeddings for maximum search precision.
Returns exact relevant text snippets, section headings, and document metadata citations.
Supports searching across technical API documentation, user manuals, and internal SOP procedures.
See how developers integrate this API endpoint into production apps.
Power instant intelligent search across thousands of technical API documentation pages and code examples.
Retrieve relevant technical documentation chunks to feed LLM prompt context for grounded chatbot responses.
Enable company employees to search internal HR policies, IT security rules, and expense guidelines instantly.
Comparing RSFlowHub's specialized API gateway vs raw multi-turn LLM prompts vs legacy regex rules.
| Evaluation Metric | RSFlowHub API | Raw LLM Prompts | Legacy Regex / Rules |
|---|---|---|---|
| Response Structure | Validated JSON envelope | Unstructured markdown or chat format | Brittle custom regex tuples |
| Setup Effort | One POST request, no prompts | Write and maintain prompts + output parsing | Fast, but rule lists miss context and sarcasm |
| Pricing Model | Flat 3 credit/request | Volatile token-based billing | High server maintenance overhead |
| Multi-Turn Hallucination | Zero (Deterministic validators) | Output format can drift between model versions | No semantic AI intelligence |
| Developer Integration | 3 lines cURL, PHP, JS, Python | Complex prompt engineering & retry loop | Fragile manual rule maintenance |
query
string
User search query or technical question to retrieve knowledge snippets for.
documents
array
Array of document objects or text chunks to search over.
top_k
integer
Number of top relevant document snippets to retrieve (Default: 5).
filter_metadata
object
Optional metadata filter key-value pairs (e.g. {"category": "billing"}).
Every endpoint returns the same envelope: success, request_id, data.result and meta.
{
"success": true,
"request_id": "req_3f9a2c7e-5b1d-4e8a-9c2f-7d6e1a0b4c8d",
"data": {
"result": {
"message": "See the API reference for the full result schema."
}
},
"meta": {
"version": "v1",
"timestamp": "2026-10-05T09:30:00Z"
}
}
Everything you need to know about integrating and using Semantic Knowledge Base Vector Search API.
Get started in minutes with free welcome credits and instant API key generation.
🐞 Found an issue? Report it and our team replies within 1 business day.
💡 Need a custom endpoint? Tell us what you need and we will scope it with you.