Embeddings Integration

Embeddings are used for vector retrieval, RAG, similarity calculation, clustering, and recommendations. Chat models and Embeddings models are typically not the same; select Model IDs explicitly marked as supporting Embeddings in the console.

Interface Information

Configuration	Value
Endpoint	https://api.corerouter.cloud/v1/embeddings
Header	Authorization: Bearer sk-...
Required Fields	model, input
Input Format	String or string array

curl Example

export COREROUTER_API_KEY="sk-xxxxxxxxxxxxxxxx"

curl https://api.corerouter.cloud/v1/embeddings \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $COREROUTER_API_KEY" \
  -d '{
    "model": "embedding-model-id",
    "input": [
      "CoreRouter is a unified AI API integration service.",
      "Embeddings can be used for search and RAG."
    ]
  }'

Python Example

from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ["COREROUTER_API_KEY"],
    base_url="https://api.corerouter.cloud/v1",
)

result = client.embeddings.create(
    model="embedding-model-id",
    input=[
        "CoreRouter is a unified AI API integration service.",
        "Embeddings can be used for search and RAG.",
    ],
)

print(len(result.data[0].embedding))

Node.js Example

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.COREROUTER_API_KEY,
  baseURL: "https://api.corerouter.cloud/v1",
});

const result = await client.embeddings.create({
  model: "embedding-model-id",
  input: [
    "CoreRouter is a unified AI API integration service.",
    "Embeddings can be used for search and RAG.",
  ],
});

console.log(result.data[0].embedding.length);

RAG Usage Recommendations

• Use the same Embeddings model for document indexing and querying.
• Save vector dimensions to avoid dimension mismatch after switching models.
• Split long documents first then vectorize; adjust split size based on business semantics.
• Model status, context, and billing information in console takes priority over model IDs in examples.

Common Issues

• Inconsistent returned dimensions: Check if different Embeddings models were mixed.
• Poor query results: Check chunking strategy, recall count, reranking logic, and original text quality.
• Returns model not found: Use Model IDs from console that support Embeddings.