Alternatives to Cohere Rerank 3

Self-hosted Decentralized Knowledge Graph node software that gives AI agents shared, verifiable memory across platforms and models.

Open-source Python framework that incrementally keeps AI agent context fresh by reprocessing only changed data.

Voyage AI embedding models and rerankers for high-precision RAG retrieval, now part of MongoDB

RustyRAG turns your documents into cited streaming answers in about 300 ms at p50, built for voice agents and live chat.

Research lab behind SID-1, an agentic search model for AI context retrieval.

Open long-context text embeddings plus agentic code-compliance and submittal review for AEC firms

Open-source RAG engine with built-in ETL, hybrid search and visual agent orchestration for enterprise AI agents

Fully managed vector database and knowledge platform: search, upsert, and rerank from one index and one API key.

Memgraph is an in-memory graph database for GraphRAG, AI agent memory, and real-time graph analytics on one engine.

VectorAI DB: run a local vector database for persistent AI agent memory on your own hardware, offline or air-gapped.

Ragie Connect syncs Google Drive, Notion, Slack, and Confluence into an agent-ready RAG context engine via API.

Context layer for AI agents: push nested JSON once, then query graph, vector, and analytics through a single API.

OpenCrab turns your documents into versioned ontology knowledge packs that Claude, ChatGPT, Cursor and Codex query through one MCP URL.

TiDB is a distributed SQL database with native vector search for AI agent memory, state, and RAG at scale.

Weaviate is an open-source AI database for vector search, hybrid search, and RAG with built-in embeddings and agent memory.

Moss delivers in-process semantic search with sub-10ms retrieval for voice agents, copilots and on-device AI.

Run LLM inference, embeddings, and ML training directly inside PostgreSQL using SQL functions, or use the Korvus SDK for RAG pipelines.

Kronotop is a distributed transactional document database with built-in vector search and millions of isolated agent namespaces per cluster.

Open-source Microsoft Research RAG that builds a knowledge graph and community summaries from your documents, then queries them in four modes.

Open ColPali visual document retrieval models and the ViDoRe V3 enterprise benchmark, all free on Hugging Face.

GPU-native database and memory layer that lets AI agents query multimodal data without leaving the GPU.

Embeddable RAG for Elixir and Phoenix — vector search, GraphRAG, and agentic retrieval inside your existing Ecto/Postgres repo.

InterLock is a proxy that authorizes, redacts, and audits every request your AI agents make to production databases, files, and APIs.

Open-source context retrieval layer that syncs your apps and databases so AI agents query live business data in natural language.

MongoDB Atlas keeps your vector embeddings in the same collection as your operational data, so RAG skips the sync pipeline.

Self-hosted RAG search engine for private document search — BM25 plus hybrid retrieval across 34 file formats.

RAGatouille brings ColBERT late-interaction retrieval into any Python RAG pipeline in a few lines of code.

AI-native graph database running vector similarity and OpenCypher traversal in one query for GraphRAG pipelines.

Fully managed vector database on Milvus for RAG, semantic search, and billion-scale AI applications.

Chroma is open-source search infrastructure for AI that keeps vectors, indexes and metadata on cheap object storage.