# rom1504/clip-retrieval

Easily compute clip embeddings and build a clip retrieval system with them

Repository: https://github.com/rom1504/clip-retrieval
Canonical: https://ross.abutalabs.com/products/clip-retrieval
Homepage: https://rom1504.github.io/clip-retrieval/
Language: Jupyter Notebook
License: MIT
License Family: permissive
Topics: semantic-search, deep-learning, multimodal, ai, clip, knn
Last push: 2026-03-28T17:21:35+00:00

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 74, release rhythm 11, longevity 100
- inputs: {"age_days": 1913, "days_push": 158, "days_rel": 383, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2795, forks 238 (observed 2026-08-28T04:07:22.328028+00:00)

## What it is
A Python toolkit for computing CLIP embeddings for images and text and building a semantic search/retrieval system on top of them. It includes components for inference, indexing, filtering, a Flask backend, and a web frontend UI.

## Use cases
- build a semantic search system over image datasets
- compute clip embeddings for millions of images
- index and query image+text embeddings with knn
- filter large image datasets by semantic similarity
- host a clip retrieval backend with a web ui
- search images by text query
- build multimodal retrieval at billion scale

## When to choose
- you need to compute CLIP embeddings at scale and search over them
- you want an end-to-end pipeline from image dataset to searchable index
- you need to filter image-text datasets semantically like LAION did

## When to avoid
- you need general-purpose vector search without CLIP models
- you want a fully managed hosted service rather than self-run tooling
- your data is text-only and you don't need multimodal embeddings

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, search-engine, vector-database, image-processing, rag
- domain: machine-learning, computer-vision, artificial-intelligence, data-science
- platform: python, cli
- tags: clip-embeddings, semantic-search, multimodal, knn-index, image-retrieval, flask-backend, search, gpu, web-server

## Member repositories
- rom1504/clip-retrieval (main) score 57

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:22.328028+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T08:14:40.819290+00:00, confidence not recorded.
  - readme: https://github.com/rom1504/clip-retrieval (fetched 2026-08-28T04:07:22.328028+00:00, sha 8096082a9c29)
  - homepage: https://rom1504.github.io/clip-retrieval/ (fetched 2026-08-29T09:54:56.005870+00:00, sha cc5abeebb0a4)
  - registry_pypi: https://pypi.org/pypi/clip-retrieval/json (fetched 2026-08-29T09:54:56.009027+00:00, sha 3def76ca46d9)
- Data as of 2026-08-30T08:39:29.467469+00:00.