MosaicML Visual Search is a scalable, open‑source framework for building image‑based retrieval systems using deep‑learning embeddings. The list below provides 20 open‑source projects that can be used as alternatives for visual search, vector indexing, and similarity retrieval.
Get targeted exposure with custom position pinning and highlighted placement.
Facebook AI Similarity Search – a library for efficient similarity search and clustering of dense vectors, widely used for image‑embedding retrieval.
High‑performance vector database supporting billions of vectors, with GPU acceleration and native integrations for image embeddings.
Open‑source vector search engine with built‑in modules for image and text embeddings, offering GraphQL/REST APIs and hybrid search capabilities.
Vector search engine optimized for high‑dimensional embeddings, providing payload filtering and real‑time updates—ideal for visual search pipelines.
Large‑scale serving engine for vector search, ranking, and recommendation, with native support for image feature vectors.
Open‑source fork of Elasticsearch that includes a k‑Nearest Neighbor plugin for fast vector similarity search on image embeddings.
Open Distro plugin that adds approximate nearest neighbor search for high‑dimensional vectors, usable for visual search workloads.
Approximate Nearest Neighbors Oh Yeah – a C++/Python library for fast, memory‑efficient similarity search, often paired with image embeddings.
Hierarchical Navigable Small World graphs implementation for high‑speed approximate nearest neighbor search, compatible with image vectors.
Non‑Metric Space Library offering a variety of ANN algorithms, suitable for large‑scale visual similarity retrieval.
Distributed vector search engine that combines vector indexing with metadata filtering, designed for image and video search.
Redis module that adds vector similarity capabilities, enabling low‑latency visual search within a familiar key‑value store.
Open‑source AI server providing image feature extraction and similarity‑search APIs, useful for building visual‑search services.
Image retrieval system that indexes visual features (color histograms, SIFT) and supports query‑by‑example search.
Open‑source visual search engine built on TensorFlow and FAISS, offering end‑to‑end indexing and query pipelines.
Toolkit within OpenCV for extracting descriptors (ORB, SIFT, etc.) and performing brute‑force or FLANN‑based visual matching.
Open‑source server that wraps CLIP models to generate image embeddings and perform similarity search via FAISS.
Community‑maintained implementation of CLIP models with training scripts, enabling custom visual embeddings for search.
Kubernetes‑native vector search engine using HNSW algorithm, supporting GPU‑accelerated indexing for image embeddings.
Microsoft's open‑source ANN library employing graph‑based indexing, optimized for high‑dimensional visual embeddings.