Softlandia-Ltd/vision-is-all-you-need
A serverless Vision RAG demo that embeds PDF pages as vectors using ColPali, stores them in Qdrant, and answers queries with GPT4o Vision.
Not currently ranked — collecting fresh signals.
star history
This project demonstrates a Vision RAG (V-RAG) architecture that eliminates traditional text chunking by converting PDF pages to images and embedding them directly using a vision-language model (ColPali). The embeddings are stored in Qdrant for semantic search. On retrieval, the user query and matched page images are passed to GPT4o Vision to generate a final answer. It runs serverless on Modal with a FastAPI backend and React frontend.
Frequently asked
- What is Softlandia-Ltd/vision-is-all-you-need?
- A serverless Vision RAG demo that embeds PDF pages as vectors using ColPali, stores them in Qdrant, and answers queries with GPT4o Vision.
- Is vision-is-all-you-need open source?
- Yes — Softlandia-Ltd/vision-is-all-you-need is open source, released under the MIT license.
- What language is vision-is-all-you-need written in?
- Softlandia-Ltd/vision-is-all-you-need is primarily written in TypeScript.
- How popular is vision-is-all-you-need?
- Softlandia-Ltd/vision-is-all-you-need has 402 stars on GitHub.
- Where can I find vision-is-all-you-need?
- Softlandia-Ltd/vision-is-all-you-need is on GitHub at https://github.com/Softlandia-Ltd/vision-is-all-you-need.