piragi
zero-config RAG for Python
python3
>>>
from piragi import Ragi
>>>
kb = Ragi("./docs")
indexing 847 chunks...
>>>
kb.retrieve("deploy to production")
[Citation(source="deploy.md", score=0.94), ...]
zero config
Works with free local models out of the box
all formats
PDF, Word, Excel, Markdown, code, URLs, images, audio
remote storage
Read from S3, GCS, Azure, HDFS with glob patterns
web crawling
Recursively crawl websites with /** syntax
auto-updates
Background refresh, queries never blocked
citations
Every answer includes scored sources
vector stores
Lance, Postgres, Qdrant, Pinecone, Supabase
retrieval
HyDE, hybrid search, cross-encoder reranking
chunking
Semantic, hierarchical, and contextual strategies
knowledge graph
Entity/relationship extraction for deeper answers
onnx
2-5x faster embeddings on CPU via ONNX Runtime
async
Non-blocking API for web frameworks
Open Playground
v0.8.1
piragi
connecting…
simulate
pip install piragi
Cells
1
Basic RAG
2
Embeddings
3
Shared Model
4
Qdrant Store
5
Chunking
6
Retrieval
7
Knowledge Graph
8
Async
9
Config
10
Ask (LLM)
11
Pipelines
12
Filter & LLM
Files
~
Use as source
LLM Config
no LLM detected
Model
Base URL
API Key
Save
Basic RAG
reset
python
Run
demo mode —
pip install piragi && piragi playground
for live execution
Data Preparation
Output
clear