FastEmbed

Health v2 · maintenance only

Adoption not part of the score

1030 stars · 145 forks observed · 2026-10-07

What it is AI-extracted, prompt v1, taxonomy v1, 2026-09-07, confidence 0.95

A Rust library for generating text and image vector embeddings and reranking documents locally using ONNX inference via ort and Hugging Face tokenizers. It supports many popular embedding models (BGE, E5, Nomic, CLIP, etc.) with synchronous usage and no Tokio dependency.

Use cases

• generate text embeddings locally in rust
• rerank search results without an api
• build a rag pipeline in rust
• compute image and text embeddings with clip
• run onnx embedding models offline
• semantic search without a server

When to choose

• you need local, offline embedding generation in Rust
• you want fast synchronous inference without Tokio
• you need both embeddings and reranking from one library
• you want multimodal text/image embeddings via CLIP

When to avoid

• you need GPU-cluster-scale training or fine-tuning
• you prefer calling a hosted embeddings API
• your stack is Python or JavaScript rather than Rust

Facets

library · maturity active

machine-learning nlp rag search-engine llm-inference machine-learning artificial-intelligence rust cross-platform embeddings onnx reranking vector-search text-embeddings natural-language-processing retrieval-augmented-generation search

Member repositories

For agents

markdown · JSON · MCP: product_card(name="anush008/fastembed-rs")