---
id: pymilvus-model
version: "0.3.2"
license: unclear
license_treatment: permissive
maintenance: aging
---

# pymilvus.model — Model components for PyMilvus, the Python SDK for Milvus

License: permissive · Maintenance: aging · Downloads: 399.4K/mo

## What it is and what it does

pymilvus-model is a library that bridges embedding and reranking models with Milvus, an open-source vector database for AI applications. It abstracts away the complexity of integrating different model providers—including commercial services like OpenAI and Voyage AI, as well as open-source options via SentenceTransformers and Hugging Face—so you can generate embeddings or rerank results without writing provider-specific integration code.

The package depends on transformers, onnxruntime, scipy, protobuf, and numpy to handle model loading, inference, and numerical operations. It is designed to work with Python 3.8 and above. The library is intended for developers building AI search or retrieval systems who need a standard way to generate or rerank embeddings.

Use it for:

- Generate embeddings from text using OpenAI, Cohere, or open-source models and store them in a vector database for semantic search.
- Rerank search results using different reranker models to improve relevance without reindexing.
- Build a retrieval-augmented generation pipeline where embeddings are created and managed through a unified interface.
- Integrate SentenceTransformers models into an application without writing custom model loading code.
- Switch between embedding providers with minimal code changes.

## Worth the install?

AI-flagged interpretation of the facts above — verify before relying.

Provides embedding and reranker model integrations for Milvus vector database, supporting OpenAI, Voyage AI, Cohere, SentenceTransformers, and Hugging Face models.

Yes, if you need a standard way to integrate embedding or reranking models with Milvus. The package has low install friction, permissive licensing, and no known vulnerabilities. However, the aging maintenance status (501 days since last release) warrants checking that its dependencies remain compatible with your environment.

## Install

pip install pymilvus-model
uv add pymilvus-model
poetry add pymilvus-model

## Installing pymilvus.model

Before you install: Low install friction with a pure-wheel distribution. Maintenance status is aging—last release was 501 days ago—but the repository remains active, suggesting ongoing support.

License in practice: Licensed under Apache Software License (permissive), allowing commercial use and modification with minimal restrictions.

Quickstart:

pip install pymilvus.model

from pymilvus_model.dense import SentenceTransformerEmbeddingFunction

embedding_fn = SentenceTransformerEmbeddingFunction()
embeddings = embedding_fn(["hello world"])

Requires Python 3.8 or above. Depending on the model provider chosen, you may need API keys or additional model files downloaded by transformers.

Verify before relying:

- Whether the package receives active maintenance independent of parent project releases or only through bundled updates.
- Current state of compatibility with the latest versions of transformers, onnxruntime, scipy, protobuf, and numpy.
- Whether pymilvus-model is included as a dependency in pymilvus or installed separately.

## Package facts

- License: not declared (permissive)
- Python support: supports_current
- Install friction: low
- Maintenance: aging
- Downloads: 399.4K/month (top 15,000 on PyPI)
- Known vulnerabilities: none known

## Tags

milvus embedding models, vector database model integration, text embedding for milvus, reranker models milvus, sentence transformers milvus, hugging face embeddings milvus, openai embeddings milvus, vector-database, embeddings, model-integration

[View on SkillFed](https://skillfed.io/packages/pymilvus-model) · [View on PyPI](https://pypi.org/project/pymilvus-model/)