Synarch

Boost your search results

Sy-rerank is the magic layer on top of your existing search to boost retrieval performance.

The magic layer

Our reranker model is designed for companies and RAG app builders to improve the relevance of your retrieval, with just a few lines of code.

import requests
import json

url = "https://api.synarch.ai/sy-rerank"
headers = {
    "Content-Type": "application/json"
}
payload = {
    "query": "best practices in RAG implementation",
    "documents": [
        {"id": "1", "text": "An introduction to Retrieval-Augmented Generation (RAG)..."},
        {"id": "2", "text": "The best open source models for RAG in 2024 are..."},
        {"id": "3", "text": "Advice for implementing Retrieval-Augmented Generation..."},
    ]
}

response = requests.post(url, headers=headers, data=json.dumps(payload))

Model features

Rerank numerous documents

Send an unlimited number of documents to the API.

Unlimited document length

Rerank documents of any length.

Universal language support

Sy-rerank supports all languages.

Lightning-fast response

Receive responses in one second.

Where we fit in your stack

Sy-rerank seamlessly integrates into the modern AI stack, working sequentially with other tools.

Query

Database / API search

Results reranking with Sy-rerank

Generation, matching, display...

Try sy-rerank simply by copying-and-pasting the code above.