Search Toolkit

Search Toolkit is a Python framework for building production-ready information retrieval (IR) systems. It provides components for ingestion, retrieval, and evaluation. It works across backends, and every component can be swapped.

LLMs are not trained on your private data. To ground their answers in your documents you need a retrieval pipeline that ingests, indexes, and searches your content. Search Toolkit gives you every building block to assemble that pipeline.

Ingestion includes:

• Multi-format extraction: PDF/DOCX/PPTX via Mistral OCR, HTML, spreadsheets, emails, plain text
• File loading: Load from the local filesystem or implement custom loaders for any source
• Flexible chunking: Character, token, markdown-aware, or separator-based splitting
• Enrichment: Enrich documents and chunks with custom metadata or LLM-generated summaries
• Indexing: Index to vector stores for semantic search

Retrieval includes:

• Multiple strategies: Vector (semantic) search with optional reranking
• Query preprocessing: Improve user queries with LLM reformulation or query extension
• Reranking: LLM reranker, cross-encoder reranker, or custom rerankers
• Semantic caching: Cache results by query similarity to skip redundant retrieval

Ingestion transforms raw documents into searchable chunks. A FileLoader reads raw bytes from a source, a DocumentExtractor converts them into a structured document, a TextSplitter divides it into chunks, an optional ChunkEnricher adds metadata, and an Embedder produces vectors for indexing into a vector store.

Retrieval finds relevant chunks for a given query. An optional query preprocessor rewrites or expands the query, a Retriever searches the vector index, and an optional Reranker re-scores the results before returning them.

Both workflows are orchestrated by high-level classes (Pipeline for ingestion, QueryEngine for retrieval) that handle component wiring and execution. Every component can be swapped: use the built-in implementations or bring your own.

Install the core package:

uv add mistralai-search-toolkit

uv add mistralai-search-toolkit

Optional extras add specialized functionality:

Install extras with the core package:

uv add "mistralai-search-toolkit[vespa]"
uv add "mistralai-search-toolkit[vespa,extractor-pymupdf]"
uv add "mistralai-search-toolkit[all]"  # Install all extras

uv add "mistralai-search-toolkit[vespa]"
uv add "mistralai-search-toolkit[vespa,extractor-pymupdf]"
uv add "mistralai-search-toolkit[all]"  # Install all extras

All packages are available on PyPI.

• Quickstart: build your first ingestion and retrieval pipeline end to end.
• Search index: set up your vector store.
• Ingestion: load, extract, chunk, enrich, and index your documents.
• Retrieval: configure vector search with optional reranking.