EmbedAnything

Use it when

• I need to create local embeddings from a PDF without building separate ingestion code.
• I need one Python pipeline to embed text files and other documented media sources.

What it solves

Not the fit when

• Choosing the best embedding model for a corpus.
• Operating a hosted vector database or search service.
• embedding-model-selection-blind
• vector-store-scale-cost

Install

pip install embed-anything

Invoke

Load `EmbeddingModel.from_pretrained_hf(...)`, then call `embed_anything.embed_file("path/to/file.pdf", embedder=model, config=config)`.

Alternatives

No reviewed alternatives recorded yet.