Lite & Super-fast re-ranking for your search & retrieval pipelines. Supports SoTA Listwise and Pairwise reranking based on LLMs and cross-encoders and more. Created by Prithivi Da, open for PRs & Collaborations.

Rapid Document Ranking, Powered by Lightweight Models.

Using FlashRank with Fast API

Robust RAG Pipeline implementation featuring Semantic Chunking (Percentile threshold), Reciprocal Rank Fusion (Hybrid Search), and Cross-Encoder Re-ranking using FlashRank. Powered by Google Gemini 1.5 Flash.

flashrank hackathon

FastAPI wrapper for https://github.com/PrithivirajDamodaran/FlashRank

No description available

A simple REST wrapper of flashrank for use as a service in n8n

Production-grade RAG System: FastAPI, Qdrant & Groq. Features Langfuse observability, Prometheus/Grafana monitoring, W&B evaluation pipeline, and FlashRank reranking.

AI Tutoring System using Retrieval Augmented Generation (RAG). Stack: Python, Streamlit, Docker, AWS, LangChain, HuggingFace, FAISS Vector DB, and FlashRank. Includes comprehensive MLflow experiment tracking.

Supplementary benchmarks for Making Legacy Knowledge Searchable with RAG

No description available

re-ranking with flashrank

No description available

No description available

No description available

Sometimes your existing search is not enough or do not priorties the right chunk/context, specially when the chunks are long orders. FlashRank ReRanker provides an effective mechanism to re-rank your existing search and prioritize the relevant context

API microservice for re-ranking using flash rank under the hood.

No description available

No description available

No description available

a benchmark for evaluating Arabic embedding models within a Retrieval-Augmented Generation (RAG) pipeline applied to Maliki jurisprudence (fiqh). The goal is to measure the quality of answers produced by a Large Language Model (llama3:8b via Ollama) depending on the choice of embedding model and the use of FlashRank Arabic re-ranking.

Production RAG (EU AI Act + arXiv) with Chroma, FlashRank rerank, OpenAI, and RAGAS metrics over time

A RAG-powered PDF chatbot built with LangChain, ChromaDB, BM25, FlashRank, and Groq's LLaMA 3.3 70B.

A high-performance, full-stack RAG engine featuring image support, FlashRank reranking, and a modern Next.js streaming chat interface

A professional Multi-Tenant RAG Platform featuring Flashrank Reranking, Metadata Filtering for data isolation, and a FastAPI/Streamlit architecture

A local-first RAG intelligence platform with HyDE, GraphRAG, and FlashRank reranking — powered by Ollama and ChromaDB. Fully private, no cloud required.

Olfactura: An Agentic RAG-based Recommendation System for Fragrances. Built with Hybrid Search (ChromaDB + FlashRank), Pydantic Tool Calling, and RAGAS-style Evaluation.

A high-performance, CPU-optimized RAG pipeline using Qdrant (HNSW), FlashRank, and Docling. Built for low-latency semantic search across thousands of local documents.

Mini RAG app with Gemini embeddings/LLM, Pinecone vector DB, FlashRank reranking, and Streamlit UI. Supports text/PDF upload, grounded answers with inline citations