Overview

SIE provides three integration paths: native SDK for full feature access, framework adapters for RAG pipelines, and OpenAI compatibility for drop-in migration.

Integration Options

Option	Features	Best For
Framework Adapters	Dense, sparse, reranking, extraction	Chroma, CrewAI, DSPy, Haystack, LangChain, LlamaIndex, Qdrant, Weaviate
Native SDK	All features, full control	Custom pipelines, advanced use cases
OpenAI Compatibility	Dense only	Migrating existing OpenAI code

---

Framework Adapters

SIE provides native packages for popular frameworks and vector stores in Python, and three in TypeScript.

Python

Framework	Package	Embeddings	Sparse	Reranking	Extraction
Chroma	sie-chroma	Yes	Yes	No	No
CrewAI	sie-crewai	No	Yes	Yes	Yes
DSPy	sie-dspy	Yes	Yes	Yes	Yes
Haystack	sie-haystack	Yes	Yes	No	No
LangChain	sie-langchain	Yes	Yes	Yes	Yes
LlamaIndex	sie-llamaindex	Yes	Yes	Yes	Yes
Qdrant	sie-qdrant	Yes	Yes	No	No
Weaviate	sie-weaviate	Yes	Yes	No	No

TypeScript

Framework	Package	Embeddings	Sparse	Reranking
Chroma	@sie/chroma	Yes	Yes	No
LangChain.js	@sie/langchain	Yes	Yes	No
LlamaIndex.ts	@sie/llamaindex	Yes	Yes	No

Use framework adapters when:

• You’re building a RAG pipeline with one of these frameworks
• You need sparse embeddings for hybrid search
• You need reranking to improve retrieval quality

---

Native SDK

The native SDK provides full access to all SIE features: dense, sparse, multi-vector embeddings, reranking, and entity extraction.

Python

pip install sie-sdk

TypeScript

pnpm add @sie/sdk

Python

from sie_sdk import SIEClient
from sie_sdk.types import Item

client = SIEClient("http://localhost:8080")

# All output types
result = client.encode(
    "BAAI/bge-m3",
    Item(text="Your text"),
    output_types=["dense", "sparse", "multivector"]
)

# Reranking
scores = client.score(
    "BAAI/bge-reranker-v2-m3",
    query=Item(text="What is AI?"),
    items=[Item(text="AI is..."), Item(text="Weather is...")]
)

# Entity extraction
entities = client.extract(
    "urchade/gliner_multi-v2.1",
    Item(text="Tim Cook leads Apple."),
    labels=["person", "organization"]
)

TypeScript

import { SIEClient } from "@sie/sdk";

const client = new SIEClient("http://localhost:8080");

// All output types
const result = await client.encode(
  "BAAI/bge-m3",
  { text: "Your text" },
  { outputTypes: ["dense", "sparse", "multivector"] }
);

// Reranking
const scores = await client.score(
  "BAAI/bge-reranker-v2-m3",
  { text: "What is AI?" },
  [{ text: "AI is..." }, { text: "Weather is..." }]
);

// Entity extraction
const entities = await client.extract(
  "urchade/gliner_multi-v2.1",
  { text: "Tim Cook leads Apple." },
  { labels: ["person", "organization"] }
);

await client.close();

Use the native SDK when:

• You’re building a custom pipeline without a framework
• You need multi-vector (ColBERT) output
• You need entity extraction
• You want fine-grained control over batching and timing

---

OpenAI Compatibility

SIE exposes /v1/embeddings matching OpenAI’s API format. Existing OpenAI code works with a URL change.

Python

from openai import OpenAI

client = OpenAI(base_url="http://localhost:8080/v1", api_key="not-needed")

response = client.embeddings.create(
    model="BAAI/bge-m3",
    input=["Your text here", "Another text"]
)

for item in response.data:
    print(f"Index {item.index}: {len(item.embedding)} dimensions")

TypeScript

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "http://localhost:8080/v1",
  apiKey: "not-needed",
});

const response = await client.embeddings.create({
  model: "BAAI/bge-m3",
  input: ["Your text here", "Another text"],
});

for (const item of response.data) {
  console.log(`Index ${item.index}: ${item.embedding.length} dimensions`);
}

Use OpenAI compatibility when:

• You have existing code using the OpenAI SDK
• You only need dense embeddings
• You want zero code changes beyond the URL

Limitations: Only dense embeddings. No sparse, multi-vector, reranking, or extraction.

---

Feature Comparison

Feature	Framework Adapters	Native SDK	OpenAI Compat
Dense embeddings	Yes	Yes	Yes
Sparse embeddings	Most	Yes	No
Multi-vector (ColBERT)	No	Yes	No
Reranking	LangChain, LlamaIndex	Yes	No
Entity extraction	LangChain, LlamaIndex, CrewAI, DSPy	Yes	No

---

What’s Next

• Chroma - embedding functions for ChromaDB
• CrewAI - sparse embeddings for hybrid search
• DSPy - embedder for DSPy retrievers
• Haystack - dense and sparse embedders
• LangChain - embeddings, reranking, and extraction for LangChain
• LlamaIndex - embeddings and reranking for LlamaIndex
• Qdrant - dense and sparse embeddings for Qdrant
• Weaviate - dense and named vectors for Weaviate
• SDK Reference - full SDK documentation