Product
The RAG stack, already assembled.
Dewey replaces the document parser, vector database, and embedding pipeline you'd otherwise assemble and maintain yourself, and ships features no other managed service offers.
Honest comparison
How Dewey stacks up
| Feature | Dewey | OpenAI File Search | Vectara | LlamaIndex |
|---|---|---|---|---|
Document parsing (PDF, DOCX, HTML, Markdown) | ||||
Chunking & embedding | ||||
Hybrid search (semantic + keyword) | partial | |||
Cited Q&A with source links | ||||
Managed infrastructure (no self-hosting) | ||||
Transparent developer pricing | partial | |||
Use any model provider (no lock-in) Dewey works with your own API key and takes no cut of generation costs. Switch providers freely. | ||||
Multi-step research (quick → exhaustive) | partial | partial | ||
Section-aware structure + lightweight section scan Documents are parsed into their natural heading hierarchy. A dedicated /sections/scan endpoint lets agents scan section summaries cheaply before deciding which chunks to retrieve — no other managed RAG service offers this. | ||||
Real-time ingestion status events | ||||
MCP server (Claude, Cursor, agents) | ||||
Time to first search result | < 5 min | < 1 hr | hours | hours |
Comparison based on publicly available documentation as of early 2026. Features and pricing change; verify with each vendor.
Why Dewey
Six things you won't find anywhere else
Sections, not just chunks
Most RAG systems split documents into fixed-size token windows and call them chunks. Dewey parses the document's actual heading hierarchy (title, section, subsection) and indexes each section as a first-class entity. Sections with generic titles like "Introduction" or "Chapter 3" automatically get AI-generated summaries so they're just as findable as sections with descriptive headings. Search results and cited answers reference the exact section by name, not an anonymous text fragment at offset 4,096.
Your AI bill, not ours
Every other managed RAG service marks up the generation costs or locks you into their proprietary models. Dewey doesn't touch your OpenAI usage. Bring your key, pay OpenAI directly, and keep full visibility over what you're spending and why.
Works where developers already work
The Dewey MCP server puts your document collections inside Claude, Cursor, and any MCP-compatible AI assistant with no custom integration required. No other RAG-as-a-service offers this. Your knowledge base, wherever your team already works.
A real API, not a framework
LlamaIndex and LangChain are powerful, but they're frameworks you deploy and maintain. Dewey is a REST API you call. No infrastructure to manage, no version upgrades to chase, no YAML pipelines to debug. Ship your app, not your RAG stack.
Zero decisions before your first result
Most RAG setups ask you to pick a vector store, configure an embedding model, and tune chunking parameters before you've seen a single result. Dewey ships with defaults that work: chunk size, overlap, hybrid search weights, embedding model. POST a file and run a query in under five minutes. Tune the knobs later, once you know what you're optimizing for.
Real-time, not polling
Document processing is async, and Dewey makes it feel instant. Server-sent events push ingestion status (pending, processing, ready) directly to your client as each file moves through the pipeline. No polling loops, no missed updates.
Ready to stop maintaining a pipeline?
Free tier, no credit card required. First search result in under five minutes.