OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
RAG & Knowledge Systems

Snappy

Snappy is an open-source document retrieval application that combines vision-language patch matching with structured OCR to find relevant regions, such as paragraphs, tables, and figures, rather than returning only whole pages.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Snappy — OpenViST project profile in RAG & Knowledge Systems.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Snappy when document search needs to return focused page regions rather than whole pages, or when visual relevance and OCR structure are both useful.

Who it is for

  • Teams building searchable document libraries.
  • Developers creating document-grounded retrieval or RAG workflows.
  • People who need to inspect which document regions contributed to a search match.

Documented uses

  • Retrieve relevant regions from documents in response to a query.
  • Use OCR-extracted text and bounding boxes alongside visual retrieval.
  • Inspect query-to-region relevance through similarity heatmaps.
Before you choose

Requirements & limits

The repository identifies Snappy as open source under the MIT license. The documented local quick start requires Docker with Compose; chat requires an OpenAI API key.

  • The retrieval method maps ColPali patch relevance to OCR bounding-box regions using spatial weighting.
  • Mean-pooling reranking is configurable and requires more compute plus a ColPali model with /patches support.
  • Optional DeepSeek OCR-2 provides extracted text, markdown, and bounding boxes, and requires an NVIDIA GPU.
  • The documented stack includes a Next.js frontend, FastAPI backend, Python, TypeScript, and Qdrant.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up Snappy locally with Docker Compose

Start the documented local services and reach the frontend and backend documentation.

Before you start

  • Docker with Compose
  • An OpenAI API key for chat

Steps

  1. Copy .env.example to .env.

  2. Set OPENAI_API_KEY in .env.

  3. Run the command below.

    Documented command
    docker compose up -d

  4. Open the frontend at http://localhost:3000 or the backend OpenAPI page at http://localhost:8000/docs.

Expected outcome

The documented local frontend and backend services are started and their stated addresses are available to open.

Guide limitations

  • The OpenAI API key is required for chat.
  • The repository instructions do not describe this as a deployment to a hosted environment.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
athrael-soju/Snappy · Official homepage title · retained in its source language
How we review sourcesSuggest a correction