# qdrant/fastembed

Fast, Accurate, Lightweight Python library to make State of the Art Embedding

Repository: https://github.com/qdrant/fastembed
Canonical: https://ross.abutalabs.com/products/fastembed
Homepage: https://qdrant.github.io/fastembed/
Language: Python
License: Apache-2.0
License Family: permissive
Topics: embeddings, openai, rag, retrieval, retrieval-augmented-generation, vector-search
Last push: 2026-08-26T01:00:21+00:00

## Health v2 (maintenance only)
Score: 83/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 64, longevity 81
- inputs: {"age_days": 1146, "days_push": 8, "days_rel": 163, "gap_med": 37, "n_releases_24m": 12}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3166, forks 234 (observed 2026-08-28T04:07:46.694928+00:00)

## What it is
FastEmbed is a lightweight Python library for generating text embeddings using ONNX Runtime instead of PyTorch, requiring no GPU and minimal dependencies. It supports a growing set of dense, multilingual, and sparse embedding models and integrates closely with Qdrant for vector search and RAG pipelines.

## Use cases
- generate text embeddings for vector search
- build a RAG pipeline without GPU or heavy PyTorch dependencies
- run embedding generation in serverless functions like AWS Lambda
- replace OpenAI Ada-002 embeddings with a local model
- embed documents and queries for retrieval with Qdrant
- compute multilingual text embeddings on CPU

## When to choose
- you need fast, accurate embeddings on CPU with a small dependency footprint
- you want to avoid downloading gigabytes of PyTorch for embedding generation
- you are deploying to serverless or constrained environments
- you use Qdrant and want built-in embedding support

## When to avoid
- you need custom model training or fine-tuning of embedding models
- you require modalities or models not in the supported list and cannot extend them
- you already have a heavy ML stack with PyTorch and GPU inference available

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, rag, llm-inference, search-engine
- domain: machine-learning
- platform: python, cross-platform, serverless
- tags: embeddings, onnx-runtime, vector-search, text-embedding, cpu-inference, qdrant, retrieval-augmented-generation, search, natural-language-processing

## Member repositories
- qdrant/fastembed (main) score 83

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:46.694928+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T07:25:10.378981+00:00, confidence not recorded.
  - readme: https://github.com/qdrant/fastembed (fetched 2026-08-28T04:07:46.694928+00:00, sha 5bba852c9bed)
  - homepage: https://qdrant.github.io/fastembed/ (fetched 2026-08-29T09:39:52.680636+00:00, sha f40719a68040)
  - registry_pypi: https://pypi.org/pypi/fastembed/json (fetched 2026-08-29T09:39:52.683136+00:00, sha dea83c537042)
- Data as of 2026-08-30T08:39:29.467469+00:00.