AlphaCorp-AI/RustyRAG
⚡ Sub-200ms RAG API built in Rust — document ingestion, Milvus vector search, Jina AI local embeddings, and LLM streaming in a single async binary. Powered by Cerebras, Groq, and more.
Implements hybrid search combining dense HNSW vectors with BM25 sparse keyword matching via Reciprocal Rank Fusion, optionally filtered through Jina's cross-encoder reranker for relevance scoring. Built on Actix-web with Rust async/await for minimal latency, plus pluggable LLM backends (Groq, Cerebras) with Server-Sent Events streaming, layout-aware PDF ingestion via Docling, and image captioning through Llama 4 Scout vision.
Stars
66
Forks
8
Language
Rust
License
MIT
Category
Last pushed
Mar 12, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/rag/AlphaCorp-AI/RustyRAG"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
raphaelmansuy/edgequake
High-performance GraphRAG inspired from LightRag written in Rust
bosun-ai/swiftide
Fast, streaming indexing, query, and agentic LLM applications in Rust
cool-japan/oxirag
A four-layer Retrieval-Augmented Generation (RAG) engine in Rust with SMT-based logic...
pixlie/PixlieAI
Please check our new project with similar targets: https://github.com/pixlie/Pixlie
itisrohit/IsoSearch
High-performance vector search engine SIMD-accelerated HNSW and LSH.