// rag · pathwaycom/serviette

serviette

serviette est un projet de RAG et connaissances open source, créé par pathwaycom (pathwaycom/serviette sur GitHub). Il compte 58 705 étoiles sur GitHub et 1 550 forks, il est publié sous la licence MIT et a été mis à jour pour la dernière fois sur GitHub le 10 novembre 2026.

A no-code RAG server that keeps answers in sync with your documents. Works with 8 vector databases. Built on Pathway.

★ 58,7 k étoiles⑂ 1,6 k forksMITmis à jour nov. 2026
▦ visuel — le README de ce projet ne contient aucune capture intégrable,
voici donc un emplacement réservé. Le guide ci-dessous est complet malgré tout.
// détails du projet

Détails du projet

Étoiles GitHub
58 705
Forks
1 550
Licence
MIT
Dernière mise à jour sur GitHub
10 novembre 2026
Type de fiche
Guide d'installation complet
Site officiel
Official site
// fonctionnement

Comment ça marche

  1. Clarify your goal

    Decide what you want from serviette and check the README and requirements to confirm it fits your setup and Python environment.

  2. Get the code

    Install from PyPI (pypi.org/project/serviette) or clone the repository from GitHub (github.com/pathwaycom/serviette) if you need the source.

  3. Install and configure

    Follow the README quickstart: get a free Pathway license key, set your embedder API key, and pick a vector database and embedder in one YAML file.

  4. Run your first task

    Work through the README's demo end to end, then point serviette at your own folder of documents.

// fonctionnalités

Pourquoi on l'adore

No-code setup

One YAML file picks the vector database and embedder — no pipeline code to write.

Live data sync

Document edits, new files and deletions are reflected in answers within seconds.

8 vector databases

Works with 8 vector databases, including Chroma, Qdrant, Milvus, Pinecone, Weaviate, pgvector, MongoDB and DuckDB.

Chat + REST API

Answers come with cited sources, over a chat UI and a REST API.

Pathway engine

Built on Pathway's streaming data engine for real-time updates.

Vector indexing

Maintains a vector index over your documents for retrieval.

// lancez-le

Lancez-le vous-même

  1. Install serviette

    Install from PyPI with pip 23.1 or newer.

    python -m pip install -U pip
    pip install serviette
  2. Set your keys

    Get a free Pathway license key and set it with your embedder API key.

    export PATHWAY_LICENSE_KEY=...
    export OPENAI_API_KEY=...
  3. Run the demo

    Start the demo and open the chat in your browser.

    serviette demo --embedder openai  # chat at http://localhost:8989
// cas d'usage

Ce que vous pouvez créer

  • Document Q&A over a folder of files that never goes stale
  • A support-team chatbot that answers from live documentation
  • A RAG server with cited answers over any of 8 vector databases
  • Prototyping a streaming RAG pipeline without building sync logic

Two smaller projects share this shelf's knowledge brief: AppFlowy-Web is the open-source AI workspace for projects, wikis and team docs, and Compiiile renders and deploys a whole folder of markdown as a site in one move. Also on this shelf: paper-qa — PaperQA2 — high-accuracy RAG for answering questions from scientific papers, with citations you can check.

// faq

Questions fréquentes

Who maintains serviette?

It is maintained by pathwaycom and contributors on GitHub, where the repository github.com/pathwaycom/serviette is the canonical home for code, issues and releases. It was previously known as llm-app.

Is serviette free to use?

Yes — it is open source under the MIT license. Read the license in the repository for what is permitted, including any conditions on redistribution or commercial use.

Does it need a paid API key?

It needs a free Pathway license key, plus your own API key for the embedder/LLM provider you choose (for example OpenAI).

How does serviette compare to alternatives?

Compared with a static-index RAG setup: serviette keeps the index continuously in sync with live documents via Pathway, instead of one-off ingestion.

How do I get started?

Install it from PyPI with pip install serviette, set your keys, and run serviette demo --embedder openai — see the README quickstart for the full steps.

// continuez l'exploration

Outils similaires

RAG et connaissances

autoflow

Graph RAG conversational knowledge base built on TiDB Serverless vector storage

★ 3,0 k étoilesApache-2.0
Voir le guide →
RAG et connaissances

SQLBot

SQLBot — DataEase's conversational analytics: ask in natural language, get SQL answers via LLMs plus RAG.

★ 6,9 k étoilesGPL-3.0
Voir le guide →
RAG et connaissances

PageIndex

PageIndex: vectorless, reasoning-based RAG that indexes documents like a human expert would

★ 39,0 k étoilesMIT
Voir le guide →
RAG et connaissances

DB-GPT

An agentic AI data assistant - chat with your databases, run SQL and code, get reports

★ 20,1 k étoilesMIT
Voir le guide →
RAG et connaissances

LEANN

The smallest vector index in the world: private RAG over everything on your laptop with 97% less storage.

★ 13,0 k étoilesMIT
Voir le guide →
RAG et connaissances

UltraRAG

A low-code MCP framework for building complex, innovative RAG pipelines — visibly, not as a black box.

★ 5,7 k étoilesApache-2.0
Voir le guide →
// alternatives

Alternatives en RAG et connaissances

Les autres projets de la catégorie RAG et connaissances avec le plus d'étoiles du catalogue LibreAI — la sélection réaliste d'alternatives si serviette ne vous convient pas.

  • MarkItDown★ 193,5 k

    MarkItDown — Microsoft tool that converts PDFs, Office files, images and more to clean Markdown for LLMs, MIT

  • Dify★ 157,8 k

    Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or…

  • ragflow★ 91,6 k

    RAGFlow is a leading open-source Retrieval-Augmented Generation (RAG) engine that fuses cutting-edge RAG with Agent capabilities to create a superior context…

  • PaddleOCR★ 90,8 k

    Lightweight OCR toolkit turning PDFs and images into structured, LLM-ready data in 100+ languages

  • MinerU★ 82,5 k

    MinerU — turns PDFs and Office documents into LLM-ready Markdown and JSON for agents and RAG pipelines, Apache-2.0 plus terms

// plus en rag

Plus en RAG et connaissances

Projets de la catégorie RAG et connaissances proches dans le catalogue, les plus proches en étoiles de serviette — continuez à explorer de part et d'autre.

  • anything-llm★ 66,7 k

    Stop renting your intelligence. Own it with AnythingLLM. Everything you need for a powerful local-first agent experience

  • private-gpt★ 57,6 k

    Complete API layer for private AI apps on local models: RAG, skills, tools, MCP and text-to-SQL

  • MinerU★ 82,5 k

    MinerU — turns PDFs and Office documents into LLM-ready Markdown and JSON for agents and RAG pipelines, Apache-2.0 plus terms

  • Open Notebook★ 41,8 k

    Self-hosted NotebookLM alternative — private research notebooks, 20+ AI providers and multi-speaker podcasts, MIT

  • PaddleOCR★ 90,8 k

    Lightweight OCR toolkit turning PDFs and images into structured, LLM-ready data in 100+ languages

  • faiss★ 41,1 k

    Meta's library for efficient similarity search and clustering of dense vectors, in C++ and Python

Obtenir serviette

Gratuit et ouvert : récupérez le code source, mettez une étoile au dépôt et commencez à créer.