Qdrant FastEmbed README · inference-research

Fields

id

sources/embed-qdrant-fastembed-github

URL

https://github.com/qdrant/fastembed

Publisher

Qdrant

Kind

repo

Accessed

2026-10-03

notes

FastEmbed README. ONNX Runtime based, explicitly no PyTorch/GB-scale torch dependency, GPU via a separate fastembed-gpu package with CUDAExecutionProvider. Supports dense, sparse (SPLADE++) and LateInteractionTextEmbedding (colbert-ir/colbertv2.0) - the only engine in this slice that ships a ColBERT multi-vector encoder path in a first-party package.

Notes

FastEmbed README. ONNX Runtime based, explicitly no PyTorch/GB-scale torch dependency, GPU via a separate fastembed-gpu package with CUDAExecutionProvider. Supports dense, sparse (SPLADE++) and LateInteractionTextEmbedding (colbert-ir/colbertv2.0) - the only engine in this slice that ships a ColBERT multi-vector encoder path in a first-party package.

Cross-references

This record references no other records.