LLM App

What is LLM App?

LLM App, now published by Pathway as serviette, is a no-code retrieval-augmented generation server for company documents. You point it at a folder, choose the vector database and embedder in a single YAML file, and start two processes, an indexer and a server. You get a chat page and a versioned REST API that answer from your files and cite their sources. Because it runs on the Pathway Live Data Framework, an edit, a new file or a deletion shows up in answers within seconds, with no manual re-indexing. It can read local folders, Google Drive, S3 or MinIO and SharePoint, and parse PDFs, Office files, scanned images, and audio or video when the matching API keys are set. Eight vector databases are supported, with DuckDB as the zero-setup default. The code is free; Pathway requires a free license key, and the only metered cost is the optional model API that writes answers.

Use cases

• Build a document Q&A server without code
• Keep a RAG index in sync with changing files
• Expose document answers through a REST API
• Search Google Drive, S3 and SharePoint content
• Answer questions from PDFs and scanned images
• Cite sources in AI answers

Who it's for

• Backend developers
• ML engineers
• Data teams
• Self-hosting teams

Why use LLM App

• Configuration lives in one YAML file, so a working document assistant needs no pipeline code.
• Answers follow document changes within seconds, which avoids stale results from a forgotten re-index.
• With the default local embedder, documents and their embeddings stay on your machine or in your database.
• Your vectors sit in a plain, documented schema in your own store, so other tools can read the same collection.
• Choosing from eight vector databases reduces lock-in to any single store.

LLM App features

• One YAML file configures the whole stack, with an interactive wizard to generate it
• Web chat page and versioned REST API on a single port
• Answers cite the documents they came from
• Live sync of edits, additions and deletions
• Eight vector databases: DuckDB, pgvector, Qdrant, Milvus, ChromaDB, Weaviate, Pinecone and MongoDB Atlas
• Sources include local files, Google Drive, S3 or MinIO, SharePoint, FTP, SFTP and WebDAV
• Parsing for PDFs, Office files, scanned images, audio and video
• Five embedder families, including free local embeddings
• Indexer and server run as separate processes, so the server scales horizontally
• Opt-in retrieval strategies such as hybrid search, reranking and query decomposition

LLM App integrations

• Pathway
• OpenAI
• Gemini
• Bedrock
• LiteLLM
• SentenceTransformers
• Google Drive
• SharePoint
• Qdrant
• Weaviate
• Milvus
• Pinecone
• pgvector
• ChromaDB
• DuckDB
• MongoDB Atlas

How to get started with LLM App

1. Install serviette with a recent pip, adding the local extra for free embeddings, and set the free PATHWAY_LICENSE_KEY.
2. Run serviette wizard to generate a YAML config, choosing a document source, vector database and embedder.
3. Optionally set OPENAI_API_KEY so the server writes answers; without it, answers quote retrieved snippets.
4. Start the stack with serviette up, which runs the indexer and server together on localhost:8989.
5. Chat in the browser or call /api/v1/retrieve, then edit files in the watched folder to see answers change.

LLM App limitations

• Needs a free Pathway license key set in PATHWAY_LICENSE_KEY
• Needs pip 23.1 or newer, since older versions can take hours to resolve dependencies
• It answers one question at a time, with no conversation memory or user accounts yet
• There is no hosted offering, so you run it yourself

LLM App pricing

LLM App is open source, so you can use and self-host it for free.

Prices can change and may vary by country or exclude tax. Check the official pricing page.

LLM App alternatives

The main LLM App alternatives listed on Debuly are LlamaIndex, LangChain, Haystack, Dify, PrivateGPT and Agentset. The table compares pricing and platforms; order is A–Z.

1. 

   Agentset

   RAG-as-a-service for developers: ingest documents, run agentic search and build chat with citations through an API and SDKs.

   Developer platformsResearch and searchAutomation and agents
   FreemiumVisit Agentset (official site)
2. 

   Dify

   Platform for building LLM apps, agents and RAG workflows on Dify Cloud or your own infrastructure, with a free Sandbox tier.

   Developer platformsAutomation and agentsChat assistants
   FreemiumVisit Dify (official site)
3. 

   Haystack

   Open-source Python framework from deepset for building RAG pipelines, AI agents and multimodal apps from modular components.

   Developer platformsCodingAutomation and agents
   Open sourceVisit Haystack (official site)
4. 

   LangChain

   Open-source framework for building LLM agents, paired with LangSmith for tracing, evals and agent deployment.

   Developer platformsAutomation and agentsCoding
   FreemiumVisit LangChain (official site)
5. 

   LlamaIndex

   Open-source framework plus the LlamaParse platform for parsing, extracting and indexing documents so LLM apps can use them.

   Developer platformsResearch and searchAutomation and agents
   FreemiumVisit LlamaIndex (official site)
6. 

   PrivateGPT

   Open-source application API for private AI that adds files, retrieval, tools and MCP on top of a self-hosted local LLM server.

   Developer platformsChat assistantsAutomation and agents
   Open sourceVisit PrivateGPT (official site)

LLM App FAQ

Sources & verification

Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.

Image credits

• Logo: from avatars.githubusercontent.com · fetched 24 Oct 2026 · Vendor trademark, used only to identify the product; removed on request
• Cover: none collected yet, so a generated tile is shown.

Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.