Skip to main content
SearchAnything Light Logo

​
What is SearchAnything?

SearchAnything is a multimodal search engine API designed for developers who need to build powerful search experiences without managing complex infrastructure. Unlike traditional search engines that only index text, SearchAnything understands the content of your files, regardless of their format.

​
Supported Modalities

Documents

Full-text search for PDFs, Word docs, and text files with precise page-level citations.

Images (OCR)

Automatically extracts and indexes text from scanned documents, screenshots, and photos.

Audio & Video

Transcribes spoken words into searchable text with timestamped locators.

Structured Data

Connect your databases and structured files for unified retrieval.

​
Why SearchAnything?

Building a RAG (Retrieval-Augmented Generation) pipeline or a search feature usually involves stitching together:
  1. OCR Service (e.g., Tesseract, AWS Textract)
  2. Transcription Service (e.g., Whisper, Deepgram)
  3. Vector Database (e.g., Pinecone, Milvus)
  4. Embedding Model (e.g., OpenAI, Cohere)
SearchAnything unifies this into a single API. You upload a file, we handle the processing pipeline, and you get a searchable index instantly.

​
Ready to build?

Quickstart Guide

Get your API key and make your first search in under 5 minutes.

API Reference

Explore our REST API endpoints for upload, search, and management.